A waste gas treatment system for urban ecological environment

Abstract

The present invention relates to the field of atmospheric pollution prevention and control technology, specifically to an exhaust gas treatment system for urban ecological environment, the system comprising: a pressure difference sensing module, a sedimentation triggering module, a spectrum analysis module, and an interface execution module. In the present invention, by real-time monitoring of the airflow velocity difference and the vertical pressure gradient, a multi-source data calibration mechanism is adopted to dynamically adjust the deflection angle of the wind field guide plate, and the liquid film disturbance is triggered in combination with the particle concentration distribution data, thereby effectively improving the particle sedimentation efficiency. Based on the gas-liquid interface vibration spectrum feature extraction and shear force data fusion analysis, the resonator amplitude parameters are accurately controlled to achieve the optimization of the liquid film dynamic stability. By real-time feedback of the liquid film contact time data, micro-eddy enhancement instructions are generated to form a closed-loop control mechanism, and the gaseous pollutant absorption efficiency and reaction uniformity are simultaneously enhanced, the pollutant removal efficiency stability is maintained under complex working conditions, and the system energy consumption and secondary pollution risk are reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of air pollution prevention and control, and in particular to a waste gas treatment system for urban ecological environment. Background Art

[0002] The field of atmospheric pollution prevention and control technology involves pollutant removal technology in industrial waste gas. The core issues include particulate matter separation, gaseous pollutant conversion and emission control. A pollution control system is constructed through technical means such as physical interception, chemical absorption, and catalytic reaction. Among them, the traditional waste gas treatment system refers to a device that uses physical filtration combined with chemical absorption to treat industrial emission gases. In response to the need for simultaneous removal of suspended particles and acidic gases in flue gas, a bag filter is usually configured to intercept solid particles, and an alkaline solution spray tower is used in parallel to achieve sulfur dioxide neutralization reaction. The treatment process includes waste gas collection pipelines, pretreatment units, multi-stage reaction chambers and terminal emission monitoring structures.

[0003] Traditional waste gas treatment systems rely on fixed physical interception and static chemical absorption modes. Bag dust collectors are not adaptable enough to fluctuations in air velocity, which can easily cause particulate matter to escape. The alkaline solution spray tower lacks a dynamic adjustment mechanism, and the gas-liquid contact time is significantly affected by flow changes, resulting in insufficient acid gas neutralization reaction. The coordination between multi-stage reaction chambers is weak, and a feedback control link based on real-time monitoring data has not been established. The overall response of the system lags behind the dynamic changes in pollutant concentration. Under non-steady-state emission conditions, the treatment efficiency is prone to attenuation, increasing the probability of terminal emissions exceeding the standard. Summary of the Invention

[0004] The purpose of the present invention is to solve the shortcomings of the prior art and to propose a waste gas treatment system for the urban ecological environment.

[0005] In order to achieve the above objectives, the present invention adopts the following technical solutions: A waste gas treatment system for urban ecological environment includes:

[0006] The pressure differential sensing module is used to obtain the airflow velocity difference and vertical pressure gradient values ​​in real time through a distributed pressure differential sensor array. The pressure variation collected within a 5-second period is input into the adaptive weighted fusion algorithm for multi-source data calibration, and the airflow correction instruction is generated and transmitted to the sedimentation trigger module.

[0007] A sedimentation trigger module is used to adjust the deflection angle of the wind field guide plate according to the airflow correction instruction, obtain PM2.5-PM10 concentration distribution data based on the laser scattering particle counter, and generate a liquid film disturbance trigger signal when the concentration gradient exceeds the preset sedimentation model threshold and transmit it to the spectrum analysis module;

[0008] A spectrum analysis module is used to collect the 0-20 Hz vibration spectrum of the gas-liquid interface through a three-axis vibration sensor, input the liquid film disturbance trigger signal and real-time shear force data into the empirical mode decomposition algorithm to extract the main frequency characteristics, generate the resonator amplitude parameters and pass them to the interface execution module;

[0009] The interface execution module is used to control the operating frequency of the piezoelectric ceramic array according to the resonator amplitude parameters, obtain the liquid film contact time data through the electromagnetic flowmeter, and generate a micro-eddy current enhancement instruction when the contact time is lower than the dynamic set value and feed it back to the pressure difference sensing module to update the sensor sampling frequency.

[0010] As a further solution of the present invention, the airflow correction instruction specifically includes the calibrated airflow velocity difference, the vertical pressure gradient value, and the dynamic compensation coefficient; the liquid film disturbance trigger signal includes the wind field steering angle parameter, the particle deposition threshold, and the turbulence suppression coefficient; the resonator amplitude parameter specifically refers to the interface resonance frequency, the shear force attenuation index, and the spectrum energy distribution value; the micro-vortex enhancement instruction includes the piezoelectric frequency modulation amount, the liquid film residence time threshold, and the vortex intensity gradient value.

[0011] As a further solution of the present invention, the dynamic setting value is determined by optimizing the experimental data on the correlation between the liquid film contact time and the exhaust gas purification efficiency by the gradient descent method, specifically satisfying the formula: ;in, Represents the dynamic setting value, Representative The contact time baseline value of the experiment, Represents the differential pressure sensor The pressure change of the sampling time, represents the pressure difference compensation coefficient, represents the total number of experiments;

[0012] The unit of the vertical pressure gradient value is Pascal / meter, which is obtained by converting the kilometer unit in the sensor raw data into the meter unit.

[0013] As a further solution of the present invention, the pressure difference sensing module includes:

[0014] The pressure differential data acquisition submodule collects the airflow velocity difference and vertical pressure gradient values ​​through a distributed pressure differential sensor array, sets the sampling frequency to obtain the original pressure differential data, divides the periodic measurement values ​​into data units according to the time window, and uses the sliding average method to perform noise reduction to obtain a stable pressure differential data set;

[0015] The multi-source calibration submodule calls the time series variation of the stable pressure difference data set, uses an adaptive weighted fusion algorithm to calculate the spatial correlation coefficient between sensor nodes, establishes a residual compensation model based on Euclidean distance, eliminates the measurement deviation between arrays by iteratively updating the weight values ​​of multiple nodes, and obtains the fused pressure difference coefficient;

[0016] The correction instruction generation submodule extracts the gradient distribution characteristics of the fused pressure difference coefficient, constructs a three-dimensional pressure difference distribution matrix, uses a dynamic interpolation algorithm to calculate the axial compensation amount, combines the preset stability threshold to generate speed correction parameters and direction adjustment parameters, and outputs the airflow correction instruction.

[0017] As a further solution of the present invention, the combination of the dynamic interpolation algorithm and the three-dimensional pressure difference distribution matrix improves the airflow correction accuracy through the space vector superposition compensation mechanism, and the compensation amount calculation satisfies ,in, Represents the axial compensation amount, Represents the gradient modulus of the three-dimensional pressure difference distribution matrix, Represents the radian value of the angle between the airflow direction and the normal of the guide plate, Represents the dynamic interpolation coefficient.

[0018] As a further solution of the present invention, the settlement triggering module includes:

[0019] The wind field control submodule analyzes the wind speed compensation coefficient and angle correction value in the airflow correction instruction, drives the servo motor shaft through the PID control algorithm, uses the Hall sensor to collect the displacement pulse signal of the guide plate, inputs the difference between the measured angle and the target angle into the proportional integral operation unit, and outputs the deflection angle value of the guide plate;

[0020] The particle monitoring submodule activates the light source excitation unit of the laser scattering device, captures the scattered light intensity distribution of particles in the 0.1μm-10μm particle size range, uses Mie scattering theory to invert the particle number concentration, establishes a time derivative calculation model for the difference between PM2.5 and PM10 concentrations, and outputs the PM concentration gradient value;

[0021] The disturbance trigger submodule inputs the PM concentration gradient value into a sliding time window comparator. When the gradient value within the window exceeds the dynamic threshold of the sedimentation model for three sampling cycles, it activates the frequency modulation circuit of the piezoelectric ceramic oscillator and outputs a liquid film disturbance trigger signal based on the linear relationship between the gradient exceeding the standard amplitude and the reference frequency.

[0022] As a further solution of the present invention, the integral time constant of the PID control algorithm is inversely proportional to the angle correction amount, specifically: ;in, represents the integration time constant, Represents the absolute value of the angle correction in the airflow correction instruction;

[0023] The dynamic threshold of the deposition model is calibrated through an aerosol deposition efficiency experiment, and the calibration data includes a mapping relationship between PM2.5 concentration gradient and liquid film disturbance frequency.

[0024] As a further solution of the present invention, the spectrum analysis module includes:

[0025] The vibration spectrum acquisition submodule uses a three-axis vibration sensor to collect 0-20Hz vibration signals from the gas-liquid interface. It uses a polynomial fitting algorithm to eliminate sensor zero-point drift, a wavelet threshold method to suppress high-frequency noise interference, and intercepts continuous vibration waveforms according to a fixed time window. It performs Euler angle coordinate conversion on the three-axis acceleration components to generate a three-dimensional spectrum data set.

[0026] The main frequency feature extraction submodule extracts the liquid film disturbance component in the Z-axis direction from the three-dimensional spectrum data set, synchronously aligns the liquid film disturbance trigger signal with the shear force sensor data, constructs a time-frequency energy distribution matrix through the Hilbert-Huang transform, separates six intrinsic mode functions using the empirical mode decomposition algorithm, calculates the correlation coefficient between each IMF component and the original signal, and selects components with correlation coefficients greater than 0.7 for spectrum reorganization to generate the main frequency feature vector;

[0027] The amplitude parameter generation submodule establishes a second-order mass-spring-damper system model based on the frequency domain distribution characteristics of the main frequency eigenvector, substitutes the eigenfrequency into the resonance equation to solve the amplitude-frequency response curve, calculates the ratio of the amplitude values ​​at multiple frequency points to the reference amplitude, determines the safe amplitude range based on the material fatigue strength threshold, and generates the resonator amplitude parameters.

[0028] The correlation coefficient of 0.7 is verified by a vibration spectrum energy distribution experiment. When the correlation coefficient is lower than 0.7, the liquid film disturbance efficiency decreases by 12%-15%.

[0029] As a further solution of the present invention, the interface execution module includes:

[0030] The frequency control submodule parses the frequency-amplitude mapping table in the resonator amplitude parameter, establishes a linear conversion equation between the piezoelectric ceramic drive voltage and the resonant frequency, collects ambient temperature sensor data to compensate for the temperature drift of the piezoelectric material, calculates the phase synchronization compensation value between the array units, and generates a frequency control instruction set;

[0031] The contact time monitoring submodule intercepts the rising edge timestamp of the square wave signal output by the electromagnetic flowmeter, calculates the time difference between adjacent rising edges as the single contact time, uses the third-order exponential smoothing method to process the time series data, and compares the smoothed contact time series with the dynamic threshold parameters in the frequency control instruction set point by point to generate a contact time deviation series;

[0032] When the contact time deviation sequence exceeds the set tolerance range three times in a row, the dynamic feedback submodule starts the eddy current intensity adjustment mechanism, calculates the proportional gain coefficient of the PID controller according to the slope of the deviation sequence, updates the sampling frequency parameter of the differential pressure sensor, and generates a micro-eddy current enhancement instruction.

[0033] As a further solution of the present invention, the linear transformation equation is ;in, represents the resonant frequency of the piezoelectric ceramic, represents the driving voltage, Represents the real-time temperature value collected by the ambient temperature sensor, Represents the base temperature of 25°C, represents the temperature drift compensation coefficient of the piezoelectric material, Represents the intrinsic capacitance parameter of the piezoelectric ceramic array.

[0034] Compared with the prior art, the advantages and positive effects of the present invention are:

[0035] In the present invention, by real-time monitoring of the air flow velocity difference and the vertical pressure gradient and adopting a multi-source data calibration mechanism, the deflection angle of the wind field guide plate is dynamically adjusted, and the liquid film disturbance is triggered in combination with the particle concentration distribution data, thereby effectively improving the particle settling efficiency. Based on the gas-liquid interface vibration spectrum feature extraction and shear force data fusion analysis, the resonator amplitude parameters are precisely controlled to achieve the optimization of the dynamic stability of the liquid film. By real-time feedback of the liquid film contact time data and generation of micro-eddy enhancement instructions, a closed-loop control mechanism is formed, which simultaneously enhances the gaseous pollutant absorption efficiency and reaction uniformity, maintains the stability of the pollutant removal efficiency under complex working conditions, and reduces the system energy consumption and the risk of secondary pollution. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 This is an overall flow chart of the exhaust gas treatment system of the present invention;

[0037] Figure 2 This is a flowchart of the pressure difference sensing module of the present invention;

[0038] Figure 3 This is a flowchart of the sedimentation trigger module of the present invention;

[0039] Figure 4 This is a flowchart of the spectrum analysis module of the present invention;

[0040] Figure 5 This is a workflow diagram of the interface execution module of the present invention. DETAILED DESCRIPTION

[0041] To make the purpose, technical solutions and advantages of the present invention clearer, the following is a detailed description of the technical solutions based on software implementation in conjunction with the system architecture diagram and embodiments. It should be understood that the specific embodiments described herein are only used to explain the technical solutions of the present invention and do not constitute a limitation on the scope of protection.

[0042] In the description of this invention, the system architecture relationships or data processing flows indicated by terms such as "layer," "module," "interface," "data flow," "client," and "server" are defined based on the architecture diagrams or flow charts corresponding to the embodiments. This expression is intended solely to clarify the logical relationships between the various elements of the technical solution and does not limit the physical deployment form. The term "plurality" encompasses two or more technical units, including but not limited to scalable elements such as multiple data nodes, processing threads, service instances, or functional components. The specific number will be determined based on the actual business scenario and requires special explanation.

[0043] See also Figure 1 and Figure 2 The present invention provides a technical solution: a waste gas treatment system for urban ecological environment, comprising:

[0044] The pressure differential sensing module is used to obtain the airflow velocity difference and vertical pressure gradient values ​​in real time through a distributed pressure differential sensor array. The pressure variation collected within a 5-second period is input into the adaptive weighted fusion algorithm for multi-source data calibration, and the airflow correction instruction is generated and transmitted to the sedimentation trigger module.

[0045] The pressure difference sensing module includes:

[0046] The pressure differential data acquisition submodule collects the airflow velocity difference and vertical pressure gradient values ​​through a distributed pressure differential sensor array, sets the sampling frequency to obtain the original pressure differential data, divides the periodic measurement values ​​into data units according to the time window, and uses the sliding average method to perform noise reduction to obtain a stable pressure differential data set;

[0047] The workflow of the pressure difference sensing module starts with the pressure difference data acquisition submodule. This submodule is deployed above the main urban traffic artery, for example, on a gantry structure 15 meters above the ground and spanning 8 lanes. A pressure difference sensor is deployed every 10 meters in the horizontal direction, forming a distributed array of 10 sensors. The sampling frequency is set to 10Hz, that is, the raw pressure difference data of all sensors is collected every 0.1 seconds. In a 5-second measurement cycle, each sensor collects 50 data points. In order to eliminate the instantaneous data jumps caused by high-speed vehicle passage or sudden gusts of wind, the 50 raw data points collected by each sensor are subjected to noise reduction processing using the sliding average method. Specifically, the time window width is set to 3 sampling points. Starting from the third data point, the value of the point is replaced by the arithmetic mean of the three data points, including the point and the two points before it. For example, if the raw pressure values ​​collected by sensor S1 at t1, t2, and t3 are 101325.1 Pa, 101328.5 Pa, and 101324.8 Pa, respectively, the stabilized pressure value at t3 is (101325.1 + 101328.5 + 101324.8) / 3 = 101326.13 Pa. This process slides over the entire 5-second time window, generating a stabilized pressure differential dataset consisting of 48 data points for each of the ten sensors.

[0048] The multi-source calibration submodule uses the time series variation of the stable differential pressure data set, calculates the spatial correlation coefficient between sensor nodes using an adaptive weighted fusion algorithm, establishes a residual compensation model based on Euclidean distance, eliminates inter-array measurement deviations by iteratively updating the weight values ​​of multiple nodes, and obtains the fused differential pressure coefficient.

[0049] The multi-source calibration submodule receives these 10 stable differential pressure data sets. To eliminate systematic measurement deviations caused by manufacturing tolerances, aging, or installation position differences, the spatial correlation coefficient between sensor nodes must be calculated and a compensation model established. First, the temporal variation at the corresponding time points in each data set is extracted, that is, the pressure difference between adjacent sampling points. For example, for sensor S1, its stable pressure values ​​at times t3 and t4 are 101326.13 Pa and 101326.53 Pa, respectively, resulting in a temporal variation of +0.4 Pa. This generates a 47-point variation sequence for each sensor. Next, the spatial correlation between the variation sequences of any two sensors (e.g., S1 and S2) is calculated. This is achieved by calculating the inverse of the Euclidean distance between two 47-dimensional vectors. The smaller the distance, the higher the correlation. Assuming the Euclidean distance between S1 and S2 is 5.2 and the Euclidean distance between S1 and S3 is 8.1, the correlation between S1 and S2 is higher than that between S1 and S3. Based on this, a residual compensation model is established. Using the sensor with the most stable historical data in the array (e.g., S5) as the benchmark, the weights of the other sensors are iteratively updated to compensate for measurement deviations. Initially, the weights are all 1.0. In the first iteration, the S1 reading is multiplied by the weight adjustment value (e.g., 0.998) calculated from its correlation coefficient with S5, and the residual error between this value and the S5 reading is calculated. This iterative process continues until the sum of the squared residual errors of all sensors relative to the benchmark sensor S5 falls below a preset convergence threshold (e.g., 0.01 Pa²). The resulting set of sensor weights becomes the final weight. The adjusted stable differential pressure data from each sensor is then weighted averaged to produce the fused differential pressure coefficient. For example, if at time t_k, the adjusted pressure gradient values ​​of S1, S2, …, S10 are 0.82Pa / m, 0.85Pa / m, …, 0.83Pa / m respectively, and their corresponding final weights are w1, w2, …, w10, then the fusion pressure difference coefficient at this moment is (0.82w1+0.85w2+…+0.83*w10) / (w1+w2+…+w10).

[0050] The correction instruction generation submodule extracts the gradient distribution characteristics of the fused pressure difference coefficient, constructs a three-dimensional pressure difference distribution matrix, uses a dynamic interpolation algorithm to calculate the axial compensation amount, combines the preset stability threshold to generate speed correction parameters and direction adjustment parameters, and outputs the airflow correction instruction;

[0051] The correction instruction generation submodule calls the fused pressure differential coefficient sequence generated in the above steps. First, a real-time three-dimensional pressure differential distribution matrix is ​​constructed using sensor grid data along the gantry's horizontal (X-axis) and vertical (Z-axis) directions. This matrix describes the spatial distribution of air pressure within the monitoring area. For example, a 2x5 sensor layout generates a 2x5 pressure differential matrix. The gradient distribution characteristics of this matrix are then extracted, specifically calculating the rate of change of pressure at each point in the matrix relative to its neighboring points. Axial compensation is calculated using a dynamic interpolation algorithm. This algorithm is not directly invoked, but rather is specifically executed as follows: For any non-boundary point P(i,j) in the matrix, the axial compensation is determined not only by the pressure value at that point but also by the pressure values ​​of its four neighboring points P(i-1,j), P(i+1,j), P(i,j-1), and P(i,j+1). The interpolated result at that point is calculated through a weighted average, resulting in a smoother and more accurate pressure differential distribution. This process is combined with a preset stability threshold, which is determined by analyzing the correlation between historical meteorological data and exhaust gas retention events. For example, the stability threshold is set to a standard deviation of airflow velocity fluctuations less than 0.2m / s. When the calculated pressure difference field may cause the airflow velocity fluctuations to exceed this threshold, the system will generate velocity correction parameters and direction adjustment parameters. Ultimately, the output airflow correction instruction is specifically a set of numerical values: the calibrated airflow velocity difference, such as 1.5m / s; the vertical pressure gradient value, such as -2.5Pa / m (calculated by calculating the vertical distance between the sensors, such as 2 meters, and the pressure difference between the upper and lower sensors, such as 5Pa, that is, 5Pa / 2m=2.5Pa / m, with the negative sign indicating that the upper air pressure is lower than the lower air pressure); and the dynamic compensation coefficient.

[0052] The combination of dynamic interpolation algorithm and three-dimensional pressure difference distribution matrix improves the airflow correction accuracy through the space vector superposition compensation mechanism, and the compensation calculation satisfies ;

[0053] in, Represents the axial compensation amount, Represents the gradient modulus of the three-dimensional pressure difference distribution matrix, Represents the radian value of the angle between the airflow direction and the normal of the guide plate, Represents the dynamic interpolation coefficient;

[0054] The airflow correction instructions specifically include the calibrated airflow velocity difference, vertical pressure gradient value, and dynamic compensation coefficient;

[0055] The unit of vertical pressure gradient value is Pascal / meter, which is calculated by converting the kilometer unit in the sensor raw data into the meter unit.

[0056] In the technical solution, the combination of dynamic interpolation algorithm and three-dimensional pressure difference distribution matrix improves the accuracy of airflow correction through space vector superposition compensation mechanism. The calculation of compensation amount follows the formula The logic of this formula is that the compensation amount of axial speed is It is directly proportional to three core factors: the intrinsic driving force of airflow disturbance (in terms of pressure gradient modulus length) The system's response sensitivity to the driving force (expressed in dynamic interpolation coefficients) ), and the effective angle of the guide plate on the airflow (in The benefit of the formula is that it integrates the pressure difference drive in fluid mechanics, the adjustment ability of system control, and the geometric effect of physical structure into a simple mathematical relationship, so that the calculation of the compensation amount has a clear physical meaning and multi-dimensional adaptive ability. Represents the final calculated airflow axial velocity compensation, in meters per second (m / s); It is a dimensionless dynamic interpolation coefficient, whose value is optimized based on real-time airflow stability and historical data; The gradient modulus represents the three-dimensional pressure difference distribution matrix, with the unit of Pascal / meter (Pa / m), which characterizes the degree and direction of the most drastic pressure change in space; Represents the angle between the current airflow direction and the normal direction of the guide plate, in radians (rad). The term reflects the effective component of the force exerted by the guide plate on the airflow.

[0057] To calculate this formula, we first need to determine the values ​​of each parameter. 1. Dynamic interpolation coefficient Determination of: The value is not fixed, but is set through regression analysis of previous experimental data. The experiment involves applying different airflow corrections under different initial airflow conditions and recording the rate of decrease in exhaust gas concentration.

[0058] Table 1 Dynamic interpolation coefficient k value calibration experimental data:

[0059] ;

[0060] As shown in Table 1, a series of experiments were conducted to establish the relationship between the degree of airflow turbulence (represented by the standard deviation of velocity) and In this specific example, assuming that the standard deviation of the currently monitored airflow velocity is 0.3m / s, which is between Experiments 2 and 3, the linear interpolation calculation is used to obtain The value is 0.05. 2. Gradient modulus of the three-dimensional pressure difference distribution matrix Acquisition: Through the multi-source calibration submodule of the pressure difference sensing module, a fused three-dimensional space pressure difference distribution matrix has been obtained. Assume that in a local coordinate system, the pressure difference gradient components in the X, Y, and Z directions are Pa / m, Pa / m, Pa / m. Then the gradient modulus is Calculated by taking the square root of the sum of the squares of the components: Pa / m. 3. Angle between airflow direction and normal direction of guide plate Acquisition of: This value is measured in real time by the wind field sensor. Assuming that the current mainstream wind direction is 30 degrees to the normal direction of the guide plate, , converted to radians is rad.

[0061] Substitute the parameter values ​​obtained above into the formula for calculation: m / s. The calculation result m / s indicates that to achieve the optimal airflow correction effect, a speed compensation of 0.116 m / s needs to be added to the current airflow speed along the axial direction. This value will be integrated into the airflow correction instruction and passed to the next module as a key component of the speed correction parameter.

[0062] The vertical pressure gradient value is expressed in Pascals per meter and is calculated as follows: First, if the elevation units in the raw data collected by the differential pressure sensor are kilometers, unit conversion is required. For example, if the vertical distance between the upper and lower sensors is recorded as 0.002 kilometers, it is converted to 2 meters. Next, the real-time pressure reading of the upper sensor is subtracted from the reading of the lower sensor to obtain the pressure difference (Pa), which is then divided by the converted vertical distance (meters). For example, if the upper reading is 101323 Pa and the lower reading is 101328 Pa, and the vertical distance is 2 meters, the vertical pressure gradient value is (101323 - 101328) / 2 = -2.5 Pa / m.

[0063] See also Figure 1 and Figure 3 , the sedimentation trigger module is used to adjust the deflection angle of the wind field guide plate according to the airflow correction instruction, obtain PM2.5-PM10 concentration distribution data based on the laser scattering particle counter, and generate a liquid film disturbance trigger signal when the concentration gradient exceeds the preset sedimentation model threshold and transmit it to the spectrum analysis module;

[0064] The settlement trigger module includes:

[0065] The wind field control submodule analyzes the wind speed compensation coefficient and angle correction value in the airflow correction instruction, drives the servo motor shaft through the PID control algorithm, uses the Hall sensor to collect the displacement pulse signal of the guide plate, inputs the difference between the measured angle and the target angle into the proportional integral operation unit, and outputs the deflection angle value of the guide plate;

[0066] The integral time constant of the PID control algorithm is inversely proportional to the angle correction amount, specifically: ;

[0067] in, represents the integration time constant, Represents the absolute value of the angle correction in the airflow correction instruction;

[0068] The sedimentation trigger module receives the airflow correction instruction from the pressure difference sensing module, which contains a set of specific parameters: {calibrated airflow velocity difference: 1.5m / s, vertical pressure gradient value: -2.5Pa / m, angle correction value : -5 degrees}. The wind farm control submodule first analyzes the angle correction value The PID control algorithm of the servo motor is activated to accurately drive the guide plate to rotate. The integral time constant of the algorithm is It is set dynamically based on the inverse proportional relationship with the angle correction amount.

[0069] In the technical solution, the integral time constant of the PID control algorithm is inversely proportional to the angle correction amount, and the specific calculation formula is: The logic of this formula is that when the angle correction required is When it is larger, the system needs faster response speed and stronger correction force to eliminate steady-state error, so a smaller integral time constant is required. To enhance the role of the integral link; on the contrary, when the correction amount is small, a larger one is needed to avoid system overshoot and oscillation. The benefit of the formula is that it establishes a dynamic adaptive relationship between the integral time constant and the angle correction value, which enables the PID controller to automatically adjust its integral response characteristics under different correction requirements, thereby ensuring the system's rapidity while taking into account its stability. Represents the integral time constant in the PID controller, in seconds (s); Represents the absolute value of the angle correction amount parsed from the airflow correction instruction, in degrees (°). The constant 0.2 is a weight coefficient calibrated by experiment, used to adjust right The constant 1 is to avoid the sensitivity of When it is 0, the denominator is 0, and ensure There is a maximum value.

[0070] To calculate this formula, we need to obtain the 1. Angle correction value Acquisition: According to the above, the wind field control submodule parses the angle correction value in the airflow correction instruction as -5 degrees. Therefore, its absolute value Substitute this value into the formula for calculation: s. The calculation result s indicates that for this control task requiring a 5-degree deflection, the PID controller's integral time constant is set to 0.5 seconds. The controller compares the target angle (e.g., current angle 20 degrees + correction - 5 degrees = 15 degrees) with the measured angle fed back by the Hall sensor (e.g., initially 20 degrees) to obtain the error e(t). The proportional, integral, and differential units calculate the control output based on their respective gain coefficients Kp, Ki, and Kd (where Ki = Kp / Ti) and the error e(t). After receiving the control signal, the servo motor begins to rotate, and the Hall sensor reports the real-time displacement of the guide plate in the form of pulses, for example, one pulse for every 0.1-degree rotation. The PID controller continuously adjusts the output voltage until the difference between the measured and target angles falls within a preset tolerance (e.g., 0.1 degrees), completing the deflection.

[0071] The particle monitoring submodule activates the light source excitation unit of the laser scattering device, captures the scattered light intensity distribution of particles in the 0.1μm-10μm particle size range, uses Mie scattering theory to invert the particle number concentration, establishes a time derivative calculation model for the difference between PM2.5 and PM10 concentrations, and outputs the PM concentration gradient value;

[0072] The particle monitoring submodule activates a laser scattering device to illuminate the air in a specific area. The built-in light source excitation unit emits a stable laser beam with a wavelength of 635nm. When particles such as PM2.5 and PM10 in the air pass through the laser beam, light scatters. Photodetectors positioned at specific angles capture the intensity distribution of this scattered light. The system is based on Mie scattering theory, which describes the scattering of electromagnetic waves by spherical particles. Using an inversion matrix, the system converts scattered light intensity data at different angles into particle number concentrations within the 0.1μm to 10μm size range. For example, at time t1, the measured PM2.5 concentration was 85.2μg / m³ and the PM10 concentration was 110.5μg / m³. The system continuously monitors at a frequency of 1Hz. At time t2 (one second after t1), the measured PM2.5 concentration was 89.7μg / m³ and the PM10 concentration was 112.0μg / m³. Subsequently, a time derivative calculation model was established for the difference between PM2.5 and PM10 concentrations, calculating the rate of change of each particle's concentration. The PM2.5 concentration gradient is (89.7-85.2) μg / m³ / 1s = 4.5 μg / m³ / s. The PM10 concentration gradient is (112.0-110.5) μg / m³ / 1s = 1.5 μg / m³ / s. The system selects the more dramatic PM2.5 concentration gradient, 4.5 μg / m³ / s, as the output PM concentration gradient.

[0073] The disturbance trigger submodule inputs the PM concentration gradient value into a sliding time window comparator. When the gradient value within the window exceeds the dynamic threshold of the sedimentation model for three sampling cycles, the frequency modulation circuit of the piezoelectric ceramic oscillator is activated. Based on the linear relationship between the gradient exceeding the standard amplitude and the reference frequency, a liquid film disturbance trigger signal is output.

[0074] The dynamic threshold of the deposition model is calibrated through aerosol deposition efficiency experiments. The calibration data includes the mapping relationship between PM2.5 concentration gradient and liquid film disturbance frequency.

[0075] The liquid film disturbance triggering signals include wind field steering angle parameters, particle deposition threshold, and turbulence suppression coefficient.

[0076] The disturbance trigger submodule receives this PM concentration gradient value (4.5 μg / m³ / s). This value is input into a sliding time window comparator. The comparator compares the PM concentration gradient values ​​(4.2, 4.6, 4.5) over three consecutive sampling periods (i.e., 3 seconds) with the deposition model's dynamic threshold. This dynamic threshold is not a fixed value but is calibrated through aerosol deposition efficiency experiments.

[0077] Table 2 Dynamic threshold calibration table of sedimentation model:

[0078] ;

[0079] As shown in Table 2, this threshold is related to factors such as ambient humidity and the current PM2.5 baseline concentration. Assuming the current humidity is 55% and the PM2.5 baseline concentration is 85 μg / m³, the table shows that the dynamic threshold is 4.0 μg / m³ / s. Since the gradient values ​​[4.2, 4.6, and 4.5] for three consecutive sampling periods have exceeded 4.0 μg / m³ / s, the trigger condition is met. At this point, the frequency modulation circuit of the piezoelectric ceramic oscillator is activated. The frequency of the output liquid film disturbance trigger signal is determined by the linear relationship between the gradient exceeding the threshold and the reference frequency. For example, if the reference frequency is 10 Hz and the exceeding threshold is (4.5 - 4.0) / 4.0 = 12.5%, the disturbance frequency is set to 10 Hz * (1 + 0.125) = 11.25 Hz. Ultimately, the generated liquid film disturbance trigger signal contains a set of parameters: {wind field steering angle parameter: 15 degrees, particle deposition threshold: 4.0 μg / m³ / s, turbulence suppression coefficient: 0.9}, and is passed to the spectrum analysis module.

[0080] See also Figure 1 and Figure 4 , spectrum analysis module, which is used to collect the 0-20Hz vibration spectrum of the gas-liquid interface through a three-axis vibration sensor, input the liquid film disturbance trigger signal and real-time shear force data into the empirical mode decomposition algorithm to extract the main frequency features, generate the resonator amplitude parameters and pass them to the interface execution module;

[0081] The spectrum analysis module includes:

[0082] The vibration spectrum acquisition submodule uses a three-axis vibration sensor to collect 0-20Hz vibration signals from the gas-liquid interface. It uses a polynomial fitting algorithm to eliminate sensor zero-point drift, a wavelet threshold method to suppress high-frequency noise interference, and intercepts continuous vibration waveforms according to a fixed time window. It performs Euler angle coordinate conversion on the three-axis acceleration components to generate a three-dimensional spectrum data set.

[0083] The core task of the spectrum analysis module is to analyze liquid film disturbances and generate precise execution parameters. This process begins with the vibration spectrum acquisition submodule, which uses a three-axis vibration sensor mounted on the support structure of the gas-liquid interface to collect vibration signals in real time with a frequency range of 0-20 Hz. The sensor operates at a sampling rate of 100 Hz and outputs acceleration data streams in the X, Y, and Z axes. To eliminate zero-point drift caused by temperature changes or long-term operation, the collected raw data is first processed using a polynomial fitting algorithm. Specifically, the most recent 1000 sampling points (10 seconds of data) are captured and fitted with a second-order polynomial y = ax² + bx + c. The fitted curve is subtracted from the raw data as the baseline. Subsequently, to suppress high-frequency electromagnetic noise interference generated by equipment such as motors and pumps, a wavelet thresholding method is used to reduce noise on the signal. The signal is decomposed into different wavelet scales, and a threshold (e.g., calculated based on a general thresholding rule) is set for the coefficients representing high-frequency details. Coefficients below the threshold are reset to zero to reconstruct the signal. The processed continuous vibration waveform is captured according to a fixed time window (for example, 2.56 seconds, or 256 data points). Because the sensor's installation posture may not be consistent with the absolute coordinate system, the three-axis acceleration components must be converted to Euler angles to unify them in the horizontal (X', Y') and vertical (Z') directions. This generates a three-dimensional spectrum dataset that includes the vertical vibration component along the Z' axis.

[0084] The main frequency feature extraction submodule extracts the liquid film disturbance component in the Z-axis direction from the three-dimensional spectrum data set, synchronizes the liquid film disturbance trigger signal with the shear force sensor data, constructs the time-frequency energy distribution matrix through the Hilbert-Huang transform, and uses the empirical mode decomposition algorithm to separate the six intrinsic mode functions. The correlation coefficient between each IMF component and the original signal is calculated, and the components with correlation coefficients greater than 0.7 are selected for spectrum reorganization to generate the main frequency feature vector.

[0085] The correlation coefficient of 0.7 is verified by the vibration spectrum energy distribution experiment. When the correlation coefficient is lower than 0.7, the liquid film perturbation efficiency decreases by 12%-15%;

[0086] The dominant frequency feature extraction submodule focuses on analyzing the liquid film disturbance component in the Z'-axis direction of the aforementioned dataset. It first synchronizes the timestamps of the received liquid film disturbance trigger signal (including the trigger timestamp) with the data collected by the shear force sensor (another auxiliary sensor) to ensure that the analyzed vibration data strictly corresponds to the triggering event. Next, the Hilbert-Huang Transform is used to construct the time-frequency energy distribution matrix of the vibration signal, which clearly demonstrates the temporal characteristics of the signal's frequency. The core step is to use the empirical mode decomposition (EMD) algorithm to adaptively decompose the complex Z'-axis vibration signal into a series of intrinsic mode functions (IMFs). The signal is decomposed into six IMF components, each representing an intrinsic vibration mode. To identify the frequency components most relevant to the effective liquid film disturbance, the correlation coefficient between each IMF component and the original Z'-axis signal is calculated. The calculation here does not directly call the model, but rather performs the Pearson correlation coefficient calculation: each IMF component series and the original signal series are treated as two variables, their covariance is calculated, and then divided by the product of their standard deviations. For example, the calculated correlation coefficients for IMF1 to IMF6 are: [0.91, 0.82, 0.53, 0.75, 0.31, 0.15].

[0087] The system sets a correlation coefficient threshold of 0.7. This value was determined based on a series of experimental validations. In these experiments, the vibration source was actively controlled to generate liquid film disturbances with different frequency combinations, while simultaneously measuring the removal rate of harmful substances (such as NOx) from the exhaust gas. The decomposed IMF components were individually eliminated or combined, and the changes in the removal rate were observed. Experimental data showed that when the IMF components used to reconstruct the signal included components with correlation coefficients below 0.7, the system's energy was dissipated in ineffective vibration modes, resulting in a 12% to 15% decrease in the final liquid film disturbance efficiency compared to using only components with coefficients greater than 0.7. Therefore, 0.7 was established as the critical value for screening effective modes. Based on this criterion, IMF1 (0.91), IMF2 (0.82), and IMF4 (0.75) were screened out. Subsequently, these three IMF components were spectrally recombined in the frequency domain. By superimposing their spectra, a dominant spectral distribution was formed, with the energy of this distribution primarily concentrated in a few specific frequencies, which are the dominant frequency eigenvectors.

[0088] The amplitude parameter generation submodule establishes a second-order mass-spring-damper system model based on the frequency domain distribution characteristics of the main frequency eigenvector. It substitutes the eigenfrequency into the resonance equation to solve the amplitude-frequency response curve, calculates the ratio of the amplitude values ​​at multiple frequency points to the reference amplitude, and determines the safe amplitude range based on the material fatigue strength threshold to generate the resonator amplitude parameters.

[0089] The resonator amplitude parameters specifically refer to the interface resonance frequency, the shear force attenuation index, and the spectrum energy distribution value.

[0090] The amplitude parameter generation submodule performs subsequent calculations based on the dominant frequency eigenvector. First, a second-order mass-spring-damper system model is established to describe the vibration behavior of the gas-liquid interface. Its physical parameters (mass m, spring constant k, and damping coefficient c) are determined based on pre-set parameters such as the density, surface tension, and viscosity of the liquid (e.g., the absorbent). The dominant frequency in the dominant frequency eigenvector (for example, 11.5 Hz, as determined by peak detection of the reconstructed spectrum) is substituted into the model's resonance equation. Solving the equation yields the system's amplitude-frequency response curve, which describes the amplitude response at different drive frequencies. At the eigenfrequency of 11.5 Hz, the amplitude required to achieve the desired shear force is calculated (for example, 0.4 mm). Furthermore, to ensure long-term operational reliability, the fatigue strength of the piezoelectric ceramic material must be considered. The safe amplitude range is determined based on the material handbook and accelerated aging tests, for example, between 0 and 0.6 mm. Since the calculated amplitude of 0.4 mm falls within this safe range, it is adopted. Ultimately, the generated resonator amplitude parameter is not a single value, but a set of interrelated parameters, specifically: interface resonance frequency, 11.5Hz; shear force attenuation index, 0.85 (a coefficient calibrated according to experimental data to describe the nonlinear relationship between amplitude and shear force); spectral energy distribution value, 92% (indicates the percentage of energy in the main frequency eigenvector to the total energy of the original signal), and is transmitted to downstream modules.

[0091] See also Figure 1 and Figure 5 , the interface execution module is used to control the operating frequency of the piezoelectric ceramic array according to the resonator amplitude parameter, obtain the liquid film contact time data through the electromagnetic flowmeter, and generate a micro-eddy current enhancement instruction when the contact time is lower than the dynamic setting value and feed it back to the pressure difference sensing module to update the sensor sampling frequency;

[0092] The interface execution module includes:

[0093] The frequency control submodule parses the frequency-amplitude mapping table in the resonator amplitude parameters, establishes a linear conversion equation between the piezoelectric ceramic drive voltage and the resonant frequency, collects data from the ambient temperature sensor to compensate for the temperature drift of the piezoelectric material, calculates the phase synchronization compensation value between the array elements, and generates a frequency control instruction set;

[0094] The linear transformation equation is ;

[0095] in, represents the resonant frequency of the piezoelectric ceramic, represents the driving voltage, Represents the real-time temperature value collected by the ambient temperature sensor, Represents the base temperature of 25°C, represents the temperature drift compensation coefficient of the piezoelectric material, Represents the intrinsic capacitance parameter of the piezoelectric ceramic array;

[0096] The interface execution module receives the resonator amplitude parameters from the spectrum analysis module, specifically: {interface resonance frequency: 11.5 Hz, shear attenuation exponent: 0.85, spectral energy distribution: 92%}. The frequency control submodule first analyzes this parameter set. Its core goal is to precisely control the operating state of the piezoelectric ceramic array based on the 11.5 Hz resonant frequency. To achieve this, the module establishes a linear conversion equation between drive voltage and resonant frequency.

[0097] In the technical solution, the linear conversion equation is The logic of this formula is that the resonant frequency of the piezoelectric ceramic is Mainly determined by the applied driving voltage However, its response characteristics will be affected by the inherent properties of the material and the working environment. It combines two key factors: the first is the temperature drift compensation term, because the dielectric constant and elastic modulus of the piezoelectric material will change with temperature. Deviation from base temperature and changes, thus affecting the frequency response, coefficient This effect was quantified; the second This represents the equivalent intrinsic capacitance parameter of the piezoelectric ceramic array as a whole and is its fundamental electrical characteristic. The formula is beneficial in that it not only establishes the fundamental voltage-to-frequency conversion relationship but also innovatively introduces a real-time temperature compensation mechanism. By dynamically adjusting the denominator, it offsets the impact of temperature drift on frequency output accuracy, ensuring that the target resonant frequency is accurately achieved under varying ambient temperatures.

[0098] To calculate the required driving voltage using this formula , you need to get or set all the parameter values ​​in the formula. 1. Target resonant frequency The parameter received from the upstream module is 11.5Hz. 2. Ambient temperature : The ambient temperature sensor installed near the piezoelectric ceramic array collects the temperature in real time. Assume that the current collected value is 31°C. 3. Reference temperature : Set to the standard 25℃. 4. Piezoelectric material temperature drift compensation coefficient : This coefficient is obtained through experimental calibration. The experimental process is: Place the piezoelectric ceramic in a temperature-controlled box, and under a constant driving voltage, record the change of its resonant frequency with temperature from 10℃ to 40℃. By performing linear regression analysis on the relationship between frequency offset and temperature change, the slope is obtained as .set up s / ℃. 5. Intrinsic capacitance parameters of piezoelectric ceramic array : This parameter is an inherent physical property of the piezoelectric ceramic array, measured at a reference temperature using an LCR tester, and represents a proportional constant between the voltage and the resulting displacement. s.

[0099] Reshape the formula to solve for the driving voltage : . Substitute the values ​​and calculate: The calculation results show that a driving voltage of 1.5042 V is required to be applied to the piezoelectric ceramic array. Furthermore, the frequency control submodule calculates the phase synchronization compensation value for each unit in the array to ensure that the entire array vibrates in a coordinated and consistent manner, generating a frequency control instruction set containing voltage, frequency, and phase information.

[0100] The contact time monitoring submodule intercepts the rising edge timestamp of the square wave signal output by the electromagnetic flowmeter, calculates the time difference between adjacent rising edges as the single contact time, uses the third-order exponential smoothing method to process the time series data, and compares the smoothed contact time series with the dynamic threshold parameters in the frequency control instruction set point by point to generate a contact time deviation series;

[0101] The contact time monitoring submodule uses an electromagnetic flowmeter to monitor the flow of a thin liquid film across the gas-liquid interface. The flowmeter outputs a square wave signal, with each rising edge corresponding to the arrival of a new batch of liquid film. The submodule uses a high-precision timer to capture the timestamps of two consecutive rising edges, for example, t1 = 10.52s and t2 = 12.55s, and calculates the single contact time as t2 - t1 = 2.03s. To smooth the data, the resulting contact time series is processed using a third-order exponential smoothing method. This method assigns higher weights to recent data points, making the smoothed result more responsive to real-world changes. The resulting smoothed contact time series is then compared point by point with a dynamic setpoint.

[0102] When the contact time deviation sequence exceeds the set tolerance range three times in a row, the dynamic feedback submodule activates the eddy current intensity adjustment mechanism, calculates the proportional gain coefficient of the PID controller according to the slope of the deviation sequence, updates the sampling frequency parameter of the differential pressure sensor, and generates a micro-eddy current enhancement instruction;

[0103] The micro-eddy current enhancement instructions include the piezoelectric frequency modulation amount, the liquid film residence time threshold, and the eddy current intensity gradient value;

[0104] The dynamic feedback submodule compares the smoothed measured contact time (e.g., 1.92s) with the dynamic setpoint (2.044s). The tolerance range is set to ±5%, or ±0.1022s. The current deviation is 1.92-2.044 = -0.124s, which exceeds the tolerance range. If this occurs three times consecutively, the eddy current intensity adjustment mechanism is activated. The proportional gain coefficient of the PID controller is calculated based on the slope of the contact time deviation sequence (e.g., if the deviation continues to increase, the slope is negative). This in turn updates the sampling frequency of the sensor in the differential pressure sensing module (e.g., from 10Hz to 20Hz). This ultimately generates a micro-eddy current enhancement command: {piezoelectric frequency modulation amount: +0.5Hz, liquid film residence time threshold: 2.044s, eddy current intensity gradient value: 1.2}. This command is then fed back to the differential pressure sensing module, forming a closed-loop control system.

[0105] The dynamic set value is determined by optimizing the experimental data on the correlation between the liquid film contact time and the exhaust gas purification efficiency through the gradient descent method, and specifically satisfies the formula: ;

[0106] in, Represents the dynamic setting value, Representative The contact time baseline value of the experiment, Represents the differential pressure sensor The pressure change of the sampling time, represents the pressure difference compensation coefficient, Represents the total number of experiments.

[0107] The dynamic setting value The calculation of , in the technical solution, follows the formula: The logic of this formula is that the ideal liquid film contact time It should not be fixed but should be adjusted according to the real-time airflow disturbance. It is based on a series of contact time benchmark values ​​measured under quasi-steady-state conditions. The average is based on the pressure difference, and a When the airflow disturbance is severe ( When the value is larger, it means that the physical mixing effect is enhanced, so the residence time of the liquid film can be appropriately reduced, and vice versa. The strength of this regulation is quantified as the pressure difference compensation coefficient. The formula is beneficial in that it introduces the real-time aerodynamic parameters The negatively correlated dynamic correction term enables the key process parameter, contact time, to adapt to changes in the external environment, avoiding the "one-size-fits-all" approach of using the same set of standards under all working conditions, thereby optimizing the consumption of liquid medium while ensuring purification efficiency.

[0108] For calculation , you need to set the relevant parameters:

[0109] 1. Total number of experiments : Set to 5 times, which means taking 5 groups of experimental data for optimization.

[0110] 2. Contact time benchmark value and pressure change : This is a set of paired experimental data obtained by operating the system under controlled conditions. For example: (t1=2.2s,ΔP1=0.5Pa),(t2=2.1s,ΔP2=0.8Pa),(t3=2.2s,ΔP3=0.4Pa),(t4=2.0s,ΔP4=1.2Pa),(t5=2.1s,ΔP5=0.9Pa).

[0111] 3. Pressure difference compensation coefficient : The purification efficiency is optimized by gradient descent method as a function of contact time and pressure difference. s / Pa.

[0112] Substitute the values ​​for calculation: s. The calculated dynamic setting value is 2.044 seconds.

[0113] The above examples illustrate preferred implementations of the present invention. Any equivalent adjustments to the technical solutions based on software engineering methods are within the scope of protection, including but not limited to: implementing algorithmic logic using different programming languages, service-oriented reconfiguration of functional modules, adjusting data interaction protocols, optimizing resource scheduling strategies, and other technical improvements. Any implementation scheme derived from reasonable modifications to the data processing flow, service call chain, or system architecture level that does not depart from the core technology of the present invention shall be deemed to be within the scope of protection defined by the claims of the present invention.

Claims

1. A waste gas treatment system for urban ecological environment, characterized in that: The system comprises: The pressure differential sensing module is used to obtain the airflow velocity difference and vertical pressure gradient values ​​in real time through a distributed pressure differential sensor array. The pressure variation collected within a 5-second period is input into the adaptive weighted fusion algorithm for multi-source data calibration, and the airflow correction instruction is generated and transmitted to the sedimentation trigger module. The pressure difference sensing module includes: The pressure differential data acquisition submodule collects the airflow velocity difference and vertical pressure gradient values ​​through a distributed pressure differential sensor array, sets the sampling frequency to obtain the original pressure differential data, divides the periodic measurement values ​​into data units according to the time window, and uses the sliding average method to perform noise reduction to obtain a stable pressure differential data set; The multi-source calibration submodule calls the time series variation of the stable pressure difference data set, uses an adaptive weighted fusion algorithm to calculate the spatial correlation coefficient between sensor nodes, establishes a residual compensation model based on Euclidean distance, eliminates the measurement deviation between arrays by iteratively updating the weight values ​​of multiple nodes, and obtains the fused pressure difference coefficient; The correction instruction generation submodule extracts the gradient distribution characteristics of the fused pressure difference coefficient, constructs a three-dimensional pressure difference distribution matrix, calculates the axial compensation amount using a dynamic interpolation algorithm, generates speed correction parameters and direction adjustment parameters based on a preset stability threshold, and outputs an airflow correction instruction; A sedimentation trigger module is used to adjust the deflection angle of the wind field guide plate according to the airflow correction instruction, obtain PM2.5-PM10 concentration distribution data based on the laser scattering particle counter, and generate a liquid film disturbance trigger signal when the concentration gradient exceeds the preset sedimentation model threshold and transmit it to the spectrum analysis module; The settlement trigger module includes: The wind field control submodule analyzes the wind speed compensation coefficient and angle correction value in the airflow correction instruction, drives the servo motor shaft through the PID control algorithm, uses the Hall sensor to collect the displacement pulse signal of the guide plate, inputs the difference between the measured angle and the target angle into the proportional integral operation unit, and outputs the deflection angle value of the guide plate; The particle monitoring submodule activates the light source excitation unit of the laser scattering device, captures the scattered light intensity distribution of particles in the 0.1μm-10μm particle size range, uses Mie scattering theory to invert the particle number concentration, establishes a time derivative calculation model for the difference between PM2.5 and PM10 concentrations, and outputs the PM concentration gradient value; The disturbance trigger submodule inputs the PM concentration gradient value into a sliding time window comparator. When the gradient value within the window exceeds the dynamic threshold of the sedimentation model for three sampling periods, the frequency modulation circuit of the piezoelectric ceramic oscillator is activated. Based on the linear relationship between the gradient exceeding the standard amplitude and the reference frequency, a liquid film disturbance trigger signal is output. A spectrum analysis module is used to collect the 0-20 Hz vibration spectrum of the gas-liquid interface through a three-axis vibration sensor, input the liquid film disturbance trigger signal and real-time shear force data into the empirical mode decomposition algorithm to extract the main frequency characteristics, generate the resonator amplitude parameters and pass them to the interface execution module; The interface execution module is used to control the operating frequency of the piezoelectric ceramic array according to the resonator amplitude parameters, obtain the liquid film contact time data through the electromagnetic flowmeter, and generate a micro-eddy current enhancement instruction when the contact time is lower than the dynamic set value and feed it back to the pressure difference sensing module to update the sensor sampling frequency.

2. The waste gas treatment system for urban ecological environment according to claim 1 is characterized in that: The airflow correction instruction specifically includes the calibrated airflow velocity difference, vertical pressure gradient value, and dynamic compensation coefficient; the liquid film disturbance trigger signal includes the wind field steering angle parameter, the particle deposition threshold, and the turbulence suppression coefficient; the resonator amplitude parameter specifically refers to the interface resonance frequency, the shear force attenuation index, and the spectrum energy distribution value; the micro-vortex enhancement instruction includes the piezoelectric frequency modulation amount, the liquid film residence time threshold, and the vortex intensity gradient value.

3. The waste gas treatment system for urban ecological environment according to claim 1 is characterized in that: The dynamic setting value is determined by optimizing the experimental data on the correlation between the liquid film contact time and the exhaust gas purification efficiency using the gradient descent method, and specifically satisfies the formula: ;in, Represents the dynamic setting value, Representative The contact time baseline value of the experiment, Represents the differential pressure sensor The pressure change of the sampling time, represents the pressure difference compensation coefficient, represents the total number of experiments; The unit of the vertical pressure gradient value is Pascal / meter, which is obtained by converting the kilometer unit in the sensor raw data into the meter unit.

4. The waste gas treatment system for urban ecological environment according to claim 1 is characterized in that: The combination of the dynamic interpolation algorithm and the three-dimensional pressure difference distribution matrix improves the airflow correction accuracy through the space vector superposition compensation mechanism, and the compensation amount calculation satisfies ,in, Represents the axial compensation amount, Represents the gradient modulus of the three-dimensional pressure difference distribution matrix, Represents the radian value of the angle between the airflow direction and the normal of the guide plate, Represents the dynamic interpolation coefficient.

5. The waste gas treatment system for urban ecological environment according to claim 1 is characterized in that: The integral time constant of the PID control algorithm is inversely proportional to the angle correction amount, specifically: ;in, represents the integration time constant, Represents the absolute value of the angle correction in the airflow correction instruction; The dynamic threshold of the deposition model is calibrated through an aerosol deposition efficiency experiment, and the calibration data includes a mapping relationship between PM2.5 concentration gradient and liquid film disturbance frequency.

6. The waste gas treatment system for urban ecological environment according to claim 1 is characterized in that: The spectrum analysis module includes: The vibration spectrum acquisition submodule uses a three-axis vibration sensor to collect 0-20Hz vibration signals from the gas-liquid interface. It uses a polynomial fitting algorithm to eliminate sensor zero-point drift, a wavelet threshold method to suppress high-frequency noise interference, and intercepts continuous vibration waveforms according to a fixed time window. It performs Euler angle coordinate conversion on the three-axis acceleration components to generate a three-dimensional spectrum data set. The main frequency feature extraction submodule extracts the liquid film disturbance component in the Z-axis direction from the three-dimensional spectrum data set, synchronously aligns the liquid film disturbance trigger signal with the shear force sensor data, constructs a time-frequency energy distribution matrix through the Hilbert-Huang transform, separates six intrinsic mode functions using the empirical mode decomposition algorithm, calculates the correlation coefficient between each IMF component and the original signal, and selects components with correlation coefficients greater than 0.7 for spectrum reorganization to generate the main frequency feature vector; The amplitude parameter generation submodule establishes a second-order mass-spring-damper system model based on the frequency domain distribution characteristics of the main frequency eigenvector, substitutes the eigenfrequency into the resonance equation to solve the amplitude-frequency response curve, calculates the ratio of the amplitude values ​​at multiple frequency points to the reference amplitude, determines the safe amplitude range based on the material fatigue strength threshold, and generates the resonator amplitude parameters; The correlation coefficient of 0.7 is verified by a vibration spectrum energy distribution experiment. When the correlation coefficient is lower than 0.7, the liquid film disturbance efficiency decreases by 12%-15%.

7. The waste gas treatment system for urban ecological environment according to claim 6 is characterized in that: The interface execution module includes: The frequency control submodule parses the frequency-amplitude mapping table in the resonator amplitude parameter, establishes a linear conversion equation between the piezoelectric ceramic drive voltage and the resonant frequency, collects ambient temperature sensor data to compensate for the temperature drift of the piezoelectric material, calculates the phase synchronization compensation value between the array units, and generates a frequency control instruction set; The contact time monitoring submodule intercepts the rising edge timestamp of the square wave signal output by the electromagnetic flowmeter, calculates the time difference between adjacent rising edges as the single contact time, uses the third-order exponential smoothing method to process the time series data, and compares the smoothed contact time series with the dynamic threshold parameters in the frequency control instruction set point by point to generate a contact time deviation series; When the contact time deviation sequence exceeds the set tolerance range three times in a row, the dynamic feedback submodule starts the eddy current intensity adjustment mechanism, calculates the proportional gain coefficient of the PID controller according to the slope of the deviation sequence, updates the sampling frequency parameter of the differential pressure sensor, and generates a micro-eddy current enhancement instruction.

8. The waste gas treatment system for urban ecological environment according to claim 7 is characterized in that: The linear transformation equation is ;in, represents the resonant frequency of the piezoelectric ceramic, represents the driving voltage, Represents the real-time temperature value collected by the ambient temperature sensor, Represents the base temperature of 25°C, represents the temperature drift compensation coefficient of the piezoelectric material, Represents the intrinsic capacitance parameter of the piezoelectric ceramic array.

Images