A method and system for controlling and purifying oil fume with adaptive electric field adjustment function

Abstract

This application relates to the field of kitchen fume purification, specifically disclosing a fume purification control method and system with adaptive electric field adjustment function. The method constructs a closed-loop control logic encompassing multi-dimensional signal acquisition, dynamic reference impedance modeling, spatiotemporal decoupling identification of load characteristics, ozone suppression-oriented adaptive frequency conversion pulse modulation, flashover avoidance based on energy gradient downregulation, and health diagnosis and early warning. By sensing the operating conditions in real time and intelligently adjusting the frequency, duty cycle, and voltage of the high-voltage pulse, this method collaboratively optimizes purification efficiency, ozone suppression, operational reliability, and service life throughout the entire equipment lifecycle. The system consists of a signal acquisition module, a main control module, a high-voltage power supply module, and a human-machine interaction early warning module, which work together to execute the aforementioned control method. This invention achieves coordinated control of efficient fume purification and low ozone emissions, and implements predictive maintenance based on health vectors, significantly reducing operation and maintenance costs.

Description

Technical Field

[0001] This application relates to the field of kitchen fume purification, specifically to a fume purification control method and system with adaptive electric field adjustment function. Background Technology

[0002] Electrostatic fume purification equipment uses a high-voltage electric field to charge and collect oil fume particles, making it the mainstream purification technology for commercial kitchens. However, the electric field parameters of existing equipment (such as operating voltage, pulse frequency, and duty cycle) are usually fixed values ​​or only have a limited number of manual adjustment levels, which cannot adapt to the dynamic and complex actual working conditions of commercial kitchens in real time.

[0003] The "fixed parameters, rigid control" model has significant drawbacks: During peak cooking seasons (high oil fume concentration), fixed parameters may lead to insufficient purification electric field strength, reduced purification efficiency, and failure to meet emission standards. Under low load or high humidity conditions, a fixed high-intensity electric field can cause excessive corona discharge, generating excessive ozone byproducts and accelerating the deterioration of critical components such as corona electrodes and insulators. Furthermore, factors such as accumulated oil on the plates and sudden changes in ambient temperature and humidity can easily induce flashover discharge in the electric field, which can interrupt the purification process or even damage the high-voltage power supply. Simultaneously, equipment performance naturally degrades over time, and fixed parameters cannot compensate for this aging, resulting in unstable performance throughout the equipment's lifespan. In terms of operation and maintenance, there is a lack of effective condition monitoring and early warning systems, typically requiring only post-failure repairs or periodic, indiscriminate cleaning, leading to high and untimely maintenance costs.

[0004] Therefore, there is an urgent need for an intelligent control solution that can sense the operating conditions in real time and intelligently adjust the electric field parameters, thereby synergistically optimizing purification efficiency, ozone suppression, operational reliability and service life throughout the entire life cycle of the equipment. Summary of the Invention

[0005] To address the technical problems of traditional electrostatic fume purification equipment, such as insufficient purification efficiency, excessive ozone, flashover, and accelerated equipment aging due to the inability of fixed electric field parameters to adapt to dynamically changing kitchen conditions, this application provides a fume purification control method and system with adaptive electric field adjustment function. The specific technical solution adopted is as follows:

[0006] This application proposes a method for controlling oil fume purification with adaptive electric field adjustment. This control method constructs a closed-loop control logic encompassing multi-dimensional signal acquisition, dynamic reference impedance modeling, spatiotemporal decoupling identification of load characteristics, ozone suppression-guided adaptive frequency conversion pulse modulation, flashover avoidance through energy gradient downregulation, and health vector-based diagnosis and early warning. Each step is interconnected, achieving precise, real-time, and adaptive adjustment of the electric field parameters. The specific steps are as follows:

[0007] Step 1: Multi-dimensional signal acquisition and initialization

[0008] Deploy and initialize the signal acquisition array to provide raw data for the control method. Specifically, this includes installing an integrated temperature sensor at the inlet of the purification electric field to monitor the intake air temperature. A capacitive humidity sensor is installed outside the device to collect the humidity of the kitchen environment. The output voltage of the high-voltage power supply is acquired in real time through a precision resistor voltage divider DC voltage sampling circuit. The micro-current fluctuation signal is extracted through a non-contact high-frequency current transformer and subsequent signal conditioning circuit. The sensor signal is connected to the corresponding ADC pin of the industrial-grade microprocessor to complete the calibration.

[0009] Step 2: Modeling the dynamic reference impedance throughout the entire lifecycle of the electric field

[0010] Based on pre-calibrated parameters, a dynamic reference impedance is constructed to account for the effects of equipment aging, ambient temperature and humidity, and normal oil accumulation on the plates, serving as a precise benchmark for load identification. The model is as follows: ,in, Static impedance Based on cumulative running time The logarithmic aging decay function, (Based on intake air temperature) Ambient humidity Temperature and humidity correction operator, Based on historical cumulative oil fume load The oil contamination impedance offset is predicted through online learning. The microprocessor synchronously acquires this data during each control cycle. The current dynamic reference impedance is calculated. .

[0011] Step 3: Spatiotemporal decoupling identification of load characteristics

[0012] Based on dynamic reference impedance and current micro-signal, it can accurately distinguish between three normal operating conditions (high oil fume, high humidity, clean and low load) and abnormal conditions (unexpected accumulation of oil).

[0013] The current micro-fluid signal is truncated into a sequence using a sliding window of predetermined duration. The linear trend term is removed to obtain a quasi-steady-state current fluctuation sequence. The number of data points falling into each interval in the sequence is counted, and the current fluctuation entropy of the quasi-steady-state sequence is calculated. ,in, This represents the number of steady-state fluctuation intervals of a single-segment quasi-steady-state current. This represents the probability distribution of quasi-steady-state sequence data points in each interval.

[0014] Calculate the current real-time measured impedance With dynamic reference impedance impedance offset

[0015] , construct For the horizontal axis, The vertical axis represents the two-dimensional feature space.

[0016] Based on the critical value of impedance background fluctuation Current fluctuation entropy discrimination threshold Real-time determination is performed according to the following levels: If If the value is >0, it is determined to be an abnormal accumulation of oil contaminants beyond expectations, triggering a level 2 fault alarm; if <0 and > and > If it is determined to be a high concentration of oil fume load; < and > and ≤ This is determined to be interference from a high humidity environment; if ≤ and The current fluctuation entropy is within the baseline fluctuation range of the clean low-load operating condition as calibrated in Experiment 2, and it is determined to be clean low-load.

[0017] Step S4: Ozone Suppression-Oriented Adaptive Frequency Conversion Pulse Modulation

[0018] Based on the load assessment results, and with the optimization goal of suppressing ozone formation, the pulse frequency of the high-voltage power supply is adaptively adjusted while meeting minimum purification power constraints and hardware safety boundaries. Duty cycle and DC output voltage .

[0019] Linear prediction function for ozone formation: This function matches the ozone generation pattern of corona discharge in electrostatic fume purification: excess corona initiation voltage. Larger, duty cycle The higher the pulse frequency, the more complete the corona discharge, and the higher the ozone generation; The higher the value, the lower the single-pulse injection energy under constant power, and the less ozone is generated.

[0020] Minimum purification power constraint model: This model ensures that the output power of the electric field meets the purification needs under different concentrations of oil fumes in the kitchen, and guarantees that the oil fume purification efficiency always meets the requirements.

[0021] Electric field average output power model: This formula is used to establish adjustable electrical parameters ( , , The quantization relationship between the output power and the pulse parameters is established to achieve adaptive optimization of pulse parameters under constant power constraints, while simultaneously completing real-time verification of hardware safety boundaries. All initial reference parameters are determined according to... Complete aging correction.

[0022] Each control cycle, based on the current absolute value of the impedance offset rate. calculate In satisfying Under the constraints of hardware safety limits, optimize electrical parameters according to the following priorities to minimize Prioritize increasing pulse frequency Up to the allowable upper limit; minimize the pulse duty cycle while satisfying power constraints. Minimize the output voltage while satisfying power constraints. .

[0023] Full-condition and abnormal fallback control:

[0024] High concentration of cooking fumes: Lock in rated voltage to ensure Prioritize high frequencies and reduce duty cycles.

[0025] High humidity and low oil fume: Reduced to 1.1 times the clean no-load power The maximum voltage is limited to ≤1.05 times the critical corona initiation voltage. It operates at high frequency and low duty cycle.

[0026] Clean low load: Maintain voltage at Nearby, it operates at the highest frequency and lowest duty cycle.

[0027] Unexpected oil accumulation: An alarm is triggered, using the highest frequency and lowest duty cycle, the voltage drops to 0.8 times the rated value, and guarantees... .

[0028] Step S5: Flashover avoidance based on energy gradient downregulation

[0029] Real-time monitoring and rapid suppression of flashover discharge ensure continuous equipment operation, minimizing the drop rate of high-voltage output voltage. As a criterion, when A flashover is determined to have occurred when the preset flashover threshold is exceeded. Once a flashover is detected, the flashover status flag is immediately set to 1. This is based on the current absolute value of the impedance offset rate. The pre-calibrated energy downscaler γ is invoked to reduce the target operating power to [a certain value]. ,in Power reduction is achieved by calling the corresponding pre-cached high-speed register. The duty cycle parameter enables microsecond-level response. After the voltage stabilizes and there is no flashover for 10 ms, the power is gradually restored to the target power according to the pre-calibrated safe power recovery slope. Then the flashover status flag is reset.

[0030] Step S6: Diagnosis and Early Warning Based on Health Vectors

[0031] A multidimensional health vector is constructed based on core operational data to achieve predictive maintenance.

[0032] Health assessment: Three dimensions were selected: impedance offset rate, attenuation factor, and flashover frequency, to characterize the insulator / plate condition, corona performance, and discharge uniformity, respectively, and a degradation threshold was set for each dimension.

[0033] Tiered early warning:

[0034] Level 1 Warning (Maintenance Reminder): Triggered when a single-dimensional feature reaches 80% of the degradation threshold, or when a load identification / sampling failure occurs. It is only displayed on the human-machine interface.

[0035] Level 2 warning (fault alarm): Triggered when a single / multi-dimensional feature exceeds the degradation threshold, or when oil pollution accumulates beyond expectations or flashover frequency exceeds the standard. The alarm is highlighted on the human-machine interface and uploaded to the remote platform, requiring manual reset.

[0036] The present invention also provides a system for oil fume purification and control method with adaptive electric field adjustment function, the system comprising a signal acquisition module, a main control module, a high-voltage power supply module and a human-machine interaction early warning module working together;

[0037] The signal acquisition module integrates a temperature sensor, a humidity sensor, a precision resistor voltage divider DC voltage sampling circuit, and a non-contact high-frequency current transformer. It is used to acquire in real time the intake air temperature, ambient humidity, high voltage output voltage, and current micro-fluctuation signals, and to complete signal conditioning and calibration.

[0038] The main control module is an industrial-grade microprocessor with built-in non-volatile memory and high-speed registers. It stores all calibration parameters and execution programs for dynamic reference impedance modeling, load characteristic determination, ozone suppression-oriented parameter adjustment, flashover avoidance, and multi-dimensional health vector diagnosis. It is responsible for the entire process calculation and generates control commands.

[0039] The high-voltage power supply module is a controllable industrial-grade high-voltage power supply that receives control commands from the main control module, precisely adjusts the pulse frequency, duty cycle, and DC output voltage, and has hardware protection against overcurrent, overvoltage, and overtemperature.

[0040] The human-machine interaction early warning module is a waterproof and dustproof industrial-grade touch screen that communicates with the main control module to realize the display of operating parameters, output of hierarchical early warning information, data query and remote communication. The modules work together to complete the adaptive adjustment of the electric field parameters of the fume purification and the full life cycle management of the equipment.

[0041] The system's collaborative workflow is as follows: After the system is powered on, the signal acquisition module continuously collects data; the main control module periodically processes the data, completes operating condition identification and parameter optimization calculations, and outputs control commands; the high-voltage power supply module drives the electric field to operate according to the commands; during this process, the main control module monitors flashover in real time through high-priority interrupts and performs rapid avoidance, while also conducting online health diagnostics, and issuing early warnings through the human-machine interaction module when necessary. Through the above-mentioned hardware and software collaboration, all modules jointly achieve adaptive adjustment of the electric field parameters.

[0042] The beneficial effects of this application are as follows:

[0043] The oil fume purification control method and system with adaptive electric field adjustment function provided in this application constructs a closed-loop control logic for the entire process, from multi-dimensional signal acquisition, dynamic modeling, working condition identification, intelligent modulation, flashover protection to health management. The software and hardware are deeply coordinated to adapt to the dynamic and complex actual working conditions of commercial kitchens. It fundamentally solves the core pain points of traditional electrostatic oil fume purification equipment, such as insufficient purification efficiency, ozone exceeding the standard, easy flashover discharge, accelerated equipment aging, and high operation and maintenance costs caused by fixed electric field parameters and rigid control. It achieves comprehensive synergistic optimization of purification efficiency, ozone suppression, operational reliability, equipment lifespan, and ease of operation and maintenance.

[0044] This invention achieves precise perception of different kitchen operating conditions and equipment anomalies through dynamic reference impedance modeling and spatiotemporal decoupling identification of load characteristics. This ensures that electric field parameter adjustments always align with actual needs, guaranteeing that purification efficiency meets standards under all operating conditions. An adaptive frequency conversion pulse modulation strategy, designed with ozone suppression as its guiding principle, reduces ozone byproduct generation at the root of corona discharge, balancing high-efficiency purification with environmental protection requirements. A microsecond-level flashover avoidance mechanism based on energy gradient downregulation can quickly suppress flashover discharge and maintain basic purification capabilities, significantly improving the continuity and stability of equipment operation under harsh conditions. Simultaneously, aging correction rules are incorporated into all core parameters to compensate for performance degradation throughout the equipment's lifecycle, ensuring long-term stable operation. Combined with a multi-dimensional health vector and a graded early warning mechanism, predictive maintenance of equipment status is achieved, transforming traditional passive maintenance into precise proactive protection, significantly reducing the operation and maintenance costs of commercial kitchens. Attached Figure Description

[0045] To more clearly illustrate the technical solutions and advantages in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0046] Figure 1 This is a flowchart of an oil fume purification and control method with adaptive electric field adjustment function according to this application. Detailed Implementation

[0047] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0048] Please see Figure 1 This embodiment provides a method for controlling oil fume purification with adaptive electric field adjustment. This embodiment is applied to commercial kitchen electrostatic oil fume purification equipment and industrial oil mist purification equipment. In view of the defects of existing equipment with fixed electric field parameters that cannot adapt to dynamic working conditions, such as insufficient purification efficiency, ozone exceeding the standard, flashover discharge, and accelerated equipment aging, a fully closed-loop intelligent control architecture is constructed, which includes multi-dimensional signal acquisition, dynamic reference impedance modeling, load characteristic identification, adaptive pulse modulation, flashover avoidance, and health diagnosis. This enables real-time adaptive adjustment of electric field parameters, taking into account both purification efficiency and ozone suppression.

[0049] The parameters of this invention were determined through three sets of calibration experiments matching the actual working conditions of commercial kitchens. Under the standard clean working conditions of commercial kitchen fume purification equipment (temperature 25℃, relative humidity 50%, no fume load), the experiments covered scenarios such as high fume concentration, high ambient humidity, and flashover discharge. The experimental equipment adopted the same electrostatic fume purification electric field, industrial-grade high-voltage power supply, and matching sensing components as in actual applications.

[0050] Experiment 1: Calibration of Basic Electrical Parameters and Model Constants

[0051] Experimental environment: Standard clean temperature and humidity environment, temperature and humidity controllable clean environment test chamber, rated operating condition test bench for commercial kitchen fume purification equipment.

[0052] Experimental Objective: To calibrate the dynamic reference impedance modeling throughout the entire lifecycle of the equipment, and to determine the fundamental parameters and model constants for ozone suppression modulation.

[0053] Experimental equipment: electrostatic fume purification electric field components of the same batch, industrial-grade high-voltage power supply, STM32H743IIK6 microprocessor, precision temperature and humidity sensor, high-voltage sampler, leakage current detector, and ozone concentration detector.

[0054] Calibration parameters and methods:

[0055] S1. Connect the electric field component to a high-voltage power supply. The control power supply is stepped up from 1 / 4 of the rated voltage to the rated voltage in a stepped manner, with a 5 kV interval between each voltage step. After application, maintain the voltage for 5 minutes until high-voltage output is achieved. Leakage current fluctuation is ≤ ±2%. Continuously collect the high-voltage and leakage current values ​​at this stage at a frequency of 1 time / second. Calculate the impedance at each sampling point of this step and take the average as the effective impedance. Repeat this process for each voltage step, averaging the effective impedances to obtain the static impedance. .

[0056] S2. Take 10 sample units from the same batch and run them continuously under the rated operating conditions of a commercial kitchen and an oil fume concentration of 100mg / m³~200mg / m³. Every 100 hours of cumulative operation, remeasure the static impedance under standard clean operating conditions and calculate the results for different cumulative operating times. Performance retention rate The measured data were fitted into an aging model using the least squares method. Determine the aging coefficient , It is a time scaling factor used to make the logarithmic term dimensionless, adapting to the degradation patterns of plate oil accumulation and corona electrode aging.

[0057] Simultaneous verification: Continuous operation was performed up to the upper limit of the normal cleaning cycle (90 days), monitoring the change in electric field impedance of all prototypes, and the difference between the measured real-time impedance and the theoretical intrinsic impedance was recorded. Defined as oil stain impedance offset Based on this experimental method, the impedance shift range caused by normal oil accumulation on the electrode plate during a routine cleaning cycle can be measured to be 0.12. Up to 0.18

[0058] It should be noted that the theoretical intrinsic impedance That is, the ideal impedance under clean conditions that only includes the effects of equipment aging and environmental temperature and humidity.

[0059] S3. Place the electric field assembly in a clean, temperature- and humidity-controlled environmental test chamber. Within the actual temperature and humidity operating range of a commercial kitchen (5℃~45℃, 30%~90%RH), select temperature and humidity network points at intervals of 5℃ and 10%RH, and calculate the electric field impedance at each point. and The ratio is obtained. , construct based on intake air temperature air humidity Correct the lookup table for the two-dimensional environment of the index.

[0060] S4. For the prototype high-voltage power supply, slowly and continuously increase the voltage from 50% of the estimated corona initiation voltage at a rate of 0.5 kV / min, simultaneously recording the voltage and leakage current values ​​and plotting the current-voltage curve. Record the voltage value when the current increases by ≥5% / kV with the voltage increase rate. Repeat the voltage increase three times and calculate the average to determine the critical corona initiation voltage under standard clean operating conditions. This serves as the initial baseline value for equipment aging correction.

[0061] S5. In a clean, oil-free, standard environment, and within the safe operating range of voltage, pulse duty cycle, and pulse frequency, select multiple representative parameter combinations for testing. Under each parameter combination, after the equipment has stabilized, use ultraviolet absorption to accurately detect the ozone generation at the air outlet. And record synchronously. , , and the critical corona initiation voltage under different parameter combinations A large number of experimental data points ( , , , , This paper addresses a steep-pulse power supply based on a full-bridge LLC resonant transform and a stainless steel barb-plate electric field structure, combined with a linear prediction function for ozone generation. The least squares method was used for fitting to obtain the proportionality coefficient that best matches the predicted trend with the measured data. .

[0062] Under the same conditions as S6 and S5, the average input power during stable operation of the electric field is measured by a power analyzer, and then substituted into the average output power model of the electric field to calculate the initial power conversion constant β, which is then matched with the power conversion efficiency of the actual power supply topology and plate structure.

[0063] Experiment 2: Load and Flashover Characteristic Threshold Calibration

[0064] Experimental environment: Standard cleanroom laboratory, commercial kitchen typical operating condition simulation chamber (high concentration of oil fume: oil fume concentration 50mg / m³~300mg / m³; high humidity and low oil fume: humidity 70%~90%RH, oil fume concentration ≤10mg / m³; clean low load: oil fume concentration ≤5mg / m³), electrostatic oil fume purification flashover discharge simulation test bench

[0065] Experimental objective: To calibrate the core thresholds and characteristic parameters for load characteristic identification and flashover avoidance.

[0066] Experimental equipment: Basic equipment for Experiment 1, fume generator, high-frequency data acquisition instrument, and electrical parameter regulator.

[0067] Calibration parameters: flashover detection threshold, energy down-regulation factor Safe power recovery slope, load characteristic space determination threshold

[0068] Calibration parameters and methods:

[0069] Flashover detection threshold: A gradually increasing DC voltage is applied to the electric field component to simulate flashover discharge caused by oil stains on the electrode plates and electrode spacing misalignment in a commercial kitchen. The high-voltage output voltage signal is acquired at a sampling frequency of 1MHz, and the voltage drop rate at the moment of each flashover is recorded. Repeated more than 100 times; in the field of high-voltage engineering flashover detection, the general safety factor range for industrial-grade high-voltage equipment is 1.1 to 1.5 times. This calibration uses 1.2 times to balance flashover detection sensitivity and anti-interference capability under field conditions. Finally, all measured values ​​are taken. The average value is multiplied by a safety factor of 1.2 to determine the flashover threshold.

[0070] Energy downregulation factor Typical impedance offset range for commercial kitchens ( =0.1~0.8, oil fume concentration 50mg / m³~300mg / m³, the two are positively correlated), and different values ​​were tested separately. The flashover suppression effect and purification efficiency were studied, with the goal of achieving a flashover suppression rate of ≥95% and a purification efficiency meeting the requirements of the national standard GB18483-2001. With the absolute value of impedance offset Piecewise linear mapping relationship: When the value is 0.1~0.3, Take 0.6; When the value is 0.3~0.5, Take 0.7; When the value is 0.5~0.8, Take 0.8.

[0071] Safe power recovery slope: Under different absolute values ​​of impedance offset rate, the probability of secondary flashover is tested for different power recovery slopes, and the maximum recovery slope calibration value with a secondary flashover probability of less than 1% is selected. The power recovered per second at the safe power recovery slope does not exceed 1% of the maximum allowable operating power of the equipment under the current operating conditions.

[0072] Load characteristic space judgment threshold: Under three working conditions—high concentration of oil fumes, high humidity and low oil fumes, and clean and low load—≥100,000 sets of characteristic coordinate points of fluctuation throughout the entire working period are collected for each condition. , Using feature coordinate points as input and setting the cluster number to 3, the K-means clustering algorithm is used to identify the independent distribution regions of the three operating conditions in the feature space; based on the clustering results, two types of core thresholds are calibrated: ① impedance background fluctuation critical value. ① Used to distinguish between clean low-load operating conditions and loaded operating conditions; ② Current fluctuation entropy discrimination threshold This is used to distinguish between high oil fume and high humidity operating conditions; at the same time, the 99.7% confidence interval of the current fluctuation entropy under clean low load conditions is determined as the background fluctuation interval, matching the normal fluctuation range of kitchen no-load / low-load conditions.

[0073] Experiment 3: Hardware Response Boundary and Control Parameter Calibration

[0074] Experimental environment: Standard cleanroom laboratory, commercial oil fume purification equipment hardware performance testing platform

[0075] Experimental objective: To determine the safe operating boundaries and control response parameters of the equipment hardware.

[0076] Experimental equipment: Hardware performance tester, PWM signal analyzer, anti-interference tester

[0077] Calibration parameters: minimum power adjustment step size, maximum power reduction ratio boundary, pulse frequency / duty cycle / voltage safe operating range, PWM parameter buffering rules, hardware interrupt response timing.

[0078] Calibration parameters and methods:

[0079] S1. Electrical Parameter Safe Operating Range: Based on GB4706.1-2020 and industry standards for oil fume purification equipment, the safe operating range for high-voltage power supply pulse frequency is determined to be 5kHz~60kHz, the safe operating range for pulse duty cycle is 5%~40%, and the safe operating range for DC output voltage is 5kV~35kV. Hardware overcurrent and overvoltage protection thresholds are set to the upper limits of the safe operating range, forming a hardware safety net. Simultaneously, for high humidity and low oil fume conditions, the minimum purification power threshold is preset to the rated no-load power under clean and dry conditions. This threshold is 1.1 times higher than the critical corona initiation voltage specified above. It can balance the stability of corona discharge in high humidity environments with the need to suppress ozone generation.

[0080] S2. The minimum executable step size and maximum safe downward adjustment ratio boundary for preset power adjustment serve as hardware constraints for the online operation of the equipment; wherein the minimum power adjustment step size is ≤0.02 times the rated no-load power under clean and dry conditions. The minimum power is set to be no less than 0.6 times the rated no-load power. It can suppress flashover while ensuring the basic purification capability of the electric field.

[0081] S3. Based on the computing power of the STM32H743IIK6 industrial-grade microprocessor and combined with the microsecond-level flashover response requirements of commercial kitchens, the PWM parameter caching rules of each control thread are calibrated, and the pulse duty cycle parameters corresponding to different γ values ​​are pre-cached to the microprocessor's high-speed register to achieve microsecond-level PWM parameter updates. At the same time, the hardware interrupt response timing is calibrated, and the voltage sampling sequence is mapped to the microprocessor's highest priority interrupt service routine to avoid response delays caused by software polling and to match the rapid suppression requirements of flashover discharge in the field.

[0082] I. Deployment and Initialization of Multi-Dimensional Signal Acquisition Array

[0083] This unit provides accurate raw data support for the entire process of computation.

[0084] An integrated temperature sensor is installed at the inlet of the cleanroom electric field to directly monitor the intake air temperature. The capacitive humidity sensor is located outside the device body and collects the background humidity of the kitchen air. ;

[0085] Voltage sampling employs a precision resistor divider DC voltage sampling circuit. A high-precision resistor network linearly attenuates the several kilovolt high voltage to the microprocessor ADC input range according to a preset ratio. After impedance matching by a voltage follower, the signal is sent to the microprocessor ADC pin. The microscopic characteristics of the current are extracted through a non-contact HFCT current transformer. The secondary output signal is amplified by an isolation operational amplifier NSI1400 and filtered by a 1kHz-100kHz bandpass filter before being sent to the high-speed ADC pin. An STM32H743IIK6 industrial-grade microprocessor is used to complete the debugging and calibration of all sensors and sampling circuits, ensuring the accuracy and continuity of the raw data acquisition.

[0086] II. Dynamic reference impedance modeling throughout the entire life cycle of the electric field

[0087] Based on the parameters calibrated in Experiment 1, this unit constructs a dynamic reference impedance model for the entire lifecycle of the oil fume purification electric field. It comprehensively considers the impact of equipment aging and degradation, fluctuations in kitchen temperature and humidity, and normal oil accumulation on the plates on the electric field impedance, providing an accurate impedance reference for subsequent load characteristic identification. The formula for the dynamic reference impedance model is: In the formula, The dynamic reference impedance is determined by the combined effects of equipment aging and degradation, environmental temperature and humidity fluctuations, and normal accumulation of historical oil contaminants. Based on historical cumulative oil fume load The predicted offset of oil pollution impedance.

[0088] 2.1 Online Learning of Oil Pollution Impedance Offset

[0089] It is a variable that evolves slowly throughout the equipment's entire lifecycle and can be reset after plate cleaning; the historical cumulative oil fume load. Defined as the cumulative high-pressure operating time of the equipment since the last cleaning, or after the equipment is first put into operation or after each physical cleaning and maintenance. and Set to zero, during normal equipment operation, the control method periodically performs online learning within the maintenance time window when the equipment has no oil fume load, avoiding interference from oil fume load on impedance detection. The specific steps are as follows:

[0090] The microprocessor acquires the real-time measured impedance at the current moment. Ambient humidity Intake air temperature and historical cumulative oil fume load According to the calibration of Experiment 1 Calculate the theoretical intrinsic impedance under a completely clean state. ;

[0091] Instantaneous oil sludge impedance offset measured within the oil-free load window Compared with the current historical cumulative load Data pairs are created and stored in an empirical dataset for online learning and prediction of oil spill impedance offsets. ;

[0092] Will The data is stored in the device's non-volatile memory to form an empirical dataset of "load accumulation - pollution offset". The current historical oil fume load accumulation is obtained by querying this empirical dataset using linear interpolation. Corresponding oil contamination impedance offset ; Maximum compensation value not exceeding This threshold was determined based on measured data from Experiment 1: For 10 prototype machines from the same batch, within a 90-day routine cleaning cycle, the maximum impedance shift caused by normal oil accumulation was [value missing]. After reserving a 2% safety margin for the project, set... As Maximum compensation value

[0093] 2.2 Real-time calculation of dynamic reference impedance

[0094] In each real-time control cycle of the equipment, the microprocessor calculates the dynamic reference impedance at the current moment based on comprehensive aging, temperature and humidity, and normal oil accumulation through the following steps. :

[0095] A1. Current cumulative running time of the microprocessor synchronous data acquisition device Ambient humidity Intake air temperature and historical cumulative oil fume load .

[0096] A2. Accumulated running time Substitute into the logarithmic aging model Calculate the aging degradation function value; using the intake air temperature and ambient air humidity Using bilinear interpolation, the relevant data of the four temperature and humidity network points closest to the current temperature and humidity in the two-dimensional environment correction lookup table constructed in Experiment 1 are called to calculate the environment correction operator value corresponding to the current kitchen temperature and humidity. ; based on cumulative load For indexing, query the continuously updated "load accumulation-contamination offset" empirical dataset for online learning, and obtain the oil contamination impedance offset through interpolation. ;

[0097] A3. Obtained from step A2 , and and the static impedance calibrated in Experiment 1 Substituting these values ​​into the dynamic reference impedance model, we obtain the dynamic reference impedance values ​​of the equipment at the current moment, after comprehensive aging, environmental conditions, and normal oil accumulation. .

[0098] III. Implementation of the Spatiotemporal Decoupling Identification Model for Load Characteristics

[0099] This unit is based on the micro-current fluctuation signal and dynamic reference impedance acquired by a multi-dimensional signal acquisition array. The constructed load feature identification model employs temporal decoupling to remove slowly varying interference from the current time series and extract rapidly varying fluctuation features related to the load. Spatial decoupling, through feature space construction, effectively separates the coupled operating conditions of high oil fume and high humidity. Finally, through logical transformation, the decoupled load features are mapped into executable load judgment results, overcoming the application limitations of the traditional single threshold method. This enables accurate differentiation between three normal operating conditions in commercial kitchens—high concentration of oil fume, high humidity and low oil fume, and clean and low load—and the abnormal state of unexpected oil accumulation.

[0100] 3.1 Time Decoupling of Micro-current Fluctuation Signals

[0101] The microprocessor extracts the signal sequence from the micro-fluctuation signal of the current input from the high-speed ADC pin using a sliding window with a preset duration of 50ms. The duration of the sliding window matches the fluctuation period of the kitchen oil fume concentration. The signal sequence within the window is fitted using the least squares method and its linear trend term is removed to obtain the quasi-steady-state current fluctuation sequence.

[0102] 3.2 Calculation of Current Fluctuation Entropy

[0103] To quantify the randomness of micro-fluid fluctuations in current, the information entropy of each quasi-steady-state sequence is calculated:

[0104] Determine the amplitude range of the quasi-steady-state sequence and divide it into a predetermined number of equal parts. A range;

[0105] Count the number of data points in the sequence that fall within each interval, and calculate the probability corresponding to each interval. ,in The interval index, with values ​​ranging from 1 to... ;

[0106] Calculate the entropy value of the sequence using the information entropy formula. In the formula, for From 1 to Perform a traversal and summation. It is the natural logarithm; calculated This is the current fluctuation entropy, used to characterize the randomness and disorder of current fluctuations. The higher the concentration of cooking fumes, the stronger the randomness of the charging and migration of the fume particles, and the higher the current fluctuation entropy. The larger the size, the better it aligns with the actual operating conditions of a commercial kitchen.

[0107] 3.3 Construction of Load Feature Space

[0108] The microprocessor uses dynamic reference impedance Real-time impedance measurement at the current moment Calculate impedance offset rate ,in This reflects the deviation between the real-time measured impedance of the electric field and the dynamic reference impedance. Higher concentrations of kitchen fumes and greater ambient humidity... The larger the absolute value;

[0109] Constructed with impedance offset rate The horizontal axis represents the entropy of current fluctuations. Using the two-dimensional Cartesian feature space with the vertical axis as the ordinate, the samples of the same sampling period... and The coordinate points are combined and mapped to the load feature space to achieve real-time mapping of load features and match the dynamic changes in kitchen operating conditions.

[0110] 3.4 Load Type Classification Determination

[0111] Based on the load characteristic space judgment threshold calibrated in Experiment 2, and combined with the actual judgment requirements of abnormal operation of commercial kitchen equipment, the microprocessor completes the load type classification judgment in the following levels. The judgment process terminates immediately after triggering the corresponding level, balancing judgment accuracy and real-time performance:

[0112] Level 1 (Abnormal State): A value >0 indicates an unexpected accumulation of oil contamination on the electrode plates, reflecting the real-time measured instantaneous oil contamination impedance shift. The current cumulative load has been exceeded. The corresponding oil contamination impedance offset It has no overlap with the feature spatial distribution of the three normal operating conditions, and can be directly used as the core basis for judging abnormal states, triggering a level two fault alarm.

[0113] It should be noted that the dynamic reference impedance Approved The maximum amount of normal oil buildup Compensation, if A value >0 indicates that the accumulated oil on the electrode plates has exceeded the normal range within a regular cleaning cycle, making it impossible to pass through. The compensation offsets the losses, therefore it was directly determined to be an abnormal accumulation of oil pollution beyond expectations.

[0114] Second level (high concentration of oil fume load): <0 and > and > High concentration of oil fume load, suitable for typical cooking conditions such as stir-frying, deep-frying, and grilling in the kitchen;

[0115] Level 3 (High Humidity Environment Interference): < and > and ≤ High humidity environment interference, suitable for high humidity and low oil fume working conditions such as kitchen steaming and cooking, plum rain season, and damp flue after cleaning;

[0116] Level 4 (Clean Low Load / Normal State): ≤ and The current fluctuation entropy is within the baseline fluctuation range of the clean low-load operating condition as specified in Experiment 2, and is judged to be a clean low-load / normal state, matching the no-load / low-load operating conditions of non-cooking and light cooking in the kitchen.

[0117] IV. Ozone Suppression-Oriented Adaptive Frequency Conversion Pulse Modulation Strategy

[0118] This unit takes the load type determination results and impedance offset rate quantification data, and based on the core model constants and hardware safety boundary parameters calibrated in Experiment 1 and Experiment 3, it transforms the load characteristics into PWM drive instructions executable by the high-voltage power supply. This fully matches the dual industry requirements of efficient fume purification and low ozone emissions in commercial kitchens, suppressing the generation of ozone byproducts from the root cause of corona discharge, and achieving synergistic control of efficient fume purification and low ozone emissions.

[0119] 4.1 Core Model and Parameter Construction

[0120] Based on the physical laws governing ozone generation through pulsed corona discharge and considering the industry characteristics of commercial kitchen fume purification, a linear prediction model for ozone generation is constructed to quantify ozone generation trends. Simultaneously, a minimum purification power constraint model is set to ensure that ozone suppression meets purification efficiency requirements under different kitchen fume concentrations. Furthermore, an average output power model for the electric field is defined, establishing a quantitative correspondence between adjustable electrical parameters and output power. Specifically:

[0121] Linear prediction function for ozone formation: In the formula, This represents the ozone generation rate under the current electrical parameters. The ozone generation ratio coefficient calibrated in Experiment 1; This is the real-time DC output voltage of the high-voltage power supply; The critical corona initiation voltage calibrated in Experiment 1; This refers to the percentage of the conduction time of a single pulse in a high-voltage power supply (pulse duty cycle). This is the pulse output frequency of the high-voltage power supply; this function matches the ozone generation law of corona discharge in electrostatic fume purification: excess corona initiation voltage. Larger, duty cycle The higher the pulse frequency, the more complete the corona discharge and the higher the ozone generation; under the constraint of keeping the average output power constant, the pulse frequency... The higher the concentration of the single pulse, the lower the concentration of the single pulse energy, resulting in insufficient corona plasma reaction and reduced ozone generation.

[0122] Minimum purification power constraint model: In the formula: This is the rated no-load power under clean operating conditions, and it is a fixed hardware parameter. The power gain coefficient of the oil fume load is determined according to the classic Doitsch formula in the field of oil fume purification. The absolute value of the impedance offset rate represents the current oil fume load intensity; this model ensures that the electric field output power meets the purification requirements under different oil fume concentrations in the kitchen, and guarantees that the oil fume purification efficiency always meets the requirements.

[0123] Electric field average output power model: In the formula, This represents the real-time average output power of the electric field. The power conversion constant calibrated in Experiment 1 is uniquely related to the electric field plate structure, power supply topology efficiency, and hardware specifications; this formula is used to establish adjustable electrical parameters ( , , The quantitative correspondence between the output power and the pulse parameters is determined, and the pulse parameters are adaptively optimized under the constraint of keeping the average output power constant, while the real-time verification of the hardware safety boundary is completed.

[0124] Aging correction rule: the initial baseline parameters calibrated in all experiments ( , , ) and rated no-load power All equipment operates according to [the relevant regulations] while online. Aging correction is completed to offset model deviations caused by the aging of electric field plates and high-voltage power supplies throughout their entire life cycle, ensuring the accuracy of all models throughout the entire life cycle of the equipment.

[0125] 4.2 Real-time adaptive adjustment

[0126] The real-time control cycle in this step is fixed and strictly synchronized with the load identification calculation cycle and flashover detection cycle to match the rapid changes in kitchen fume conditions. The control priority is clearly defined as: flashover emergency avoidance > routine parameter adjustment in this step, with no control timing conflicts. Within each real-time control cycle of the equipment, the microprocessor executes the following closed-loop process:

[0127] The microprocessor synchronously acquires the current cycle load type determination result, absolute value of impedance offset rate, aging decay function value, real-time DC output voltage of high voltage power supply, and flashover status flag bit; if the flashover status flag bit is in the triggered state, the parameter adjustment of this cycle is immediately paused, and the adjustment process of this step is resumed after the flashover disappears and the high voltage output voltage returns to stability, so as to avoid the flashover discharge being aggravated.

[0128] Based on the current absolute value of impedance offset The minimum purification power required for the current operating condition is calculated based on the minimum purification power constraint model. Simultaneously retrieve the linear prediction function for ozone formation and the ozone formation ratio coefficient calibrated in Experiment 1. Power conversion constant Critical corona initiation voltage The hardware safety boundary parameters calibrated in Experiment 3, combined with the real-time acquired aging degradation function values, Complete the aging correction of the reference parameters, and use the corrected parameters as the reference for adaptive adjustment in this cycle;

[0129] Optimize the adjustable electrical parameters of the high-voltage power supply based on the premise of prioritizing ozone suppression and achieving purification standards. , , ): Prioritize pulse frequency Increase the hardware's operating range to the upper limit calibrated in Experiment 3, reduce the single-pulse injection energy, and lower the estimated ozone generation. Then satisfy Under the premise of minimizing the pulse duty cycle Shorten the duration of a single discharge and further reduce Finally, also in Under the premise of minimizing the real-time DC output voltage of the high-voltage power supply Reduce excess corona induction voltage ,Will The minimum level controlled within the national standard limit;

[0130] Microprocessor verification adjustment , , To ensure all parameters remain within the safe operating range of the hardware calibrated in Experiment 3, if any parameters exceed the boundaries, they will automatically lock to the corresponding boundary values ​​to prevent hardware damage; subsequently, the microprocessor will operate according to the finally determined frequency. Configure the period register of the internal PWM timer according to the duty cycle. Configure the PWM comparator register according to the corrected voltage. Adjust the DC output reference of the high-voltage power supply; after the PWM output waveform is amplified by the isolation drive circuit, it controls the on and off of the power switching device of the high-voltage power supply, applies the corresponding high-voltage pulse to the electric field, and completes the adaptive electric field adjustment of the current cycle.

[0131] The microprocessor will use the operating parameters of the current cycle. , , Ozone generation estimates The electric field output power data is synchronously output to the health diagnosis unit for health status assessment and early warning throughout the equipment's life cycle.

[0132] 4.3 Differentiated Control and Abnormal Coverage Rules for All Operating Conditions

[0133] Unexpected oil accumulation abnormal operating condition: Lock the highest pulse frequency and lowest pulse duty cycle within the hardware safety range calibrated in Experiment 3, reduce the DC operating voltage to 0.8 times the rated operating voltage after aging correction, and strictly meet the requirements throughout the process. Minimum safe power constraints are set to prevent complete deactivation of the electric field; basic oil fume purification capacity is guaranteed; and a secondary fault alarm is triggered simultaneously to prompt on-site maintenance personnel to perform physical cleaning of the plates in a timely manner.

[0134] High-concentration oil fume load conditions: Locked-in rated DC operating voltage after aging correction, strictly meeting the requirements. Minimum purification power constraints are set to ensure that the fume purification efficiency meets national standards; the pulse frequency is prioritized to the highest value within the hardware safety range, and the pulse duty cycle is simultaneously reduced while meeting power constraints, so as to achieve coordinated control of purification compliance and ozone suppression.

[0135] High humidity and low oil fume conditions: Minimize purification power Reduce to cleanroom no-load power 1.1 times

[0136] The maximum operating voltage is limited to 1.05 times the aging-corrected critical corona initiation voltage under standard cleanroom conditions. Lock the high-frequency, low-duty-cycle parameters within the hardware safety range, while satisfying Under the constraints of [the relevant regulations], ineffective corona discharge in high humidity environments can be suppressed at its source, thereby reducing ozone generation.

[0137] Clean, low-load operating conditions: Lock the highest frequency and lowest duty cycle allowed by the hardware to maintain the operating voltage at the critical corona initiation voltage. Nearby, ozone emissions will be controlled at background levels, while the standby power consumption of equipment will be reduced.

[0138] Abnormal fallback rules: When the load identification is abnormal or the sensor sampling fails, the equipment automatically locks to the safe parameters of the clean low-load operating condition to avoid equipment loss of control and simultaneously triggers the first-level maintenance reminder; when flashover avoidance is triggered, parameter adjustment is immediately suspended, and after the flashover discharge path disappears and the voltage stabilizes for 10ms, it is gradually restored to the optimal parameters of the current operating condition to avoid secondary flashover; when parameter adjustment encounters an unsolvable constraint, it automatically locks to the rated reference parameters of the current operating condition to ensure the basic purification capacity and operational safety of the equipment.

[0139] V. Flashover Avoidance Based on Energy Gradient Downregulation

[0140] Addressing the practical problems of flashover discharge in commercial kitchen fume purification equipment, such as oil stains on the electrode plates, electrode spacing misalignment, and sudden changes in ambient temperature and humidity, this unit constructs a flashover avoidance mechanism based on energy gradient downregulation, taking advantage of the self-healing characteristics of electrostatic fume purification electric field discharge. This mechanism effectively suppresses flashover discharge while maintaining the basic purification capacity of the electric field, ensuring continuous operation of the equipment under complex kitchen conditions and avoiding interruptions in fume purification and equipment shutdown due to flashover discharge. This is suitable for application scenarios of continuous cooking and unattended operation in commercial kitchens.

[0141] 5.1 Flashover Detection Criteria and Threshold Calibration

[0142] Detection Criteria: Based on the physical mechanism of flashover discharge in the electric field of oil fume purification, a quantifiable flashover detection criterion is constructed. The core analysis idea is that when flashover discharge occurs in the electric field of oil fume purification, a transient low-resistance discharge path is formed between the two electrodes, causing a sharp and rapid drop in the high-voltage output voltage. The voltage drop rate is the most crucial and reliable physical characteristic quantity for distinguishing flashover discharge from normal load fluctuations and power supply noise. Based on this, a flashover detection criterion is constructed: the voltage drop rate of the high-voltage output voltage... As a criterion, when When the preset flashover detection threshold is exceeded, it is determined to be a flashover discharge trigger. In order to achieve a fast flashover response, the voltage sampling sequence is mapped to the highest priority interrupt service routine of the microprocessor. The flashover detection and response are achieved at the microsecond level through hardware interrupt triggering, avoiding the delay caused by software polling.

[0143] The flashover status flag is defined as a binary variable, where 0 indicates that no flashover has been triggered and 1 indicates that a flashover has been triggered. It is stored in the microprocessor's general-purpose register, and the flashover status flag is initially set to 0.

[0144] 5.2 Avoiding Core Parameters and Hardware Configuration

[0145] All core parameters for flashover avoidance were calibrated through Experiment 2, and the hardware configuration was completed based on the parameters calibrated in Experiment 3. This adapts to the flashover avoidance requirements of commercial kitchens with different absolute values ​​of impedance offset rates, ensuring a balance between flashover suppression effect and basic electric field purification capability.

[0146] Energy downregulation factor This is used to preserve the basic purification power of the electric field while suppressing flashover; Experiment 2 has established this feature. The piecewise linear mapping relationship between the absolute value of the impedance offset rate and the data is stored in the microprocessor's non-volatile memory. Different absolute values ​​of the oil fume impedance offset rate are matched with different... Calibration values ​​ensure basic purification efficiency under high oil fume conditions;

[0147] Safe power recovery slope: The maximum power recovery slope with a secondary flashover probability of less than 1%, calibrated in Experiment 2, is stored in the microprocessor control program to avoid secondary flashover discharge while quickly restoring the normal operation of the equipment.

[0148] Hardware response configuration: Based on the PWM parameter caching rules calibrated in Experiment 3, different... The pulse duty cycle parameter corresponding to the value is pre-buffered into the microprocessor's high-speed register to achieve microsecond-level updates of the PWM parameters; the minimum power adjustment step size and maximum power reduction ratio boundary calibrated in Experiment 3 serve as hardware constraints for online flashover avoidance, ensuring the accuracy and safety of power adjustment.

[0149] 5.3 Real-time flashover monitoring and avoidance execution process

[0150] After completing the flashover detection criteria, full parameter calibration, and hardware response configuration, the microprocessor performs real-time flashover monitoring and avoidance execution through the following steps to ensure stable operation of the equipment under complex kitchen conditions:

[0151] F1. In each control cycle, the microprocessor synchronously acquires the high-voltage output voltage signal collected by the voltage sampling circuit and calculates the absolute value of the impedance offset rate. This characterizes the current kitchen fume load intensity and the current electric field target operating power determined based on minimum purification power constraints and ozone suppression optimization. ;

[0152] F2. Based on the real-time acquired high-voltage output voltage signal, calculate the voltage drop rate at the current moment. The result is compared with the flashover detection threshold calibrated in Experiment 2 to determine whether a flashover discharge has been triggered.

[0153] like ≤Flashover detection threshold: Keep the flashover status flag at 0 and maintain optimal modulation parameters. It runs while continuously monitoring the voltage drop rate via the highest priority hardware interrupt;

[0154] like >Flashover detection threshold: Immediately set the flashover status flag to 1 and execute the F3 flashover avoidance action;

[0155] F3, Flashover Avoidance Action Execution:

[0156] Based on the absolute value of the current kitchen fume impedance offset rate , call The mapping relationship with the absolute value of impedance offset is used to obtain the corresponding energy down-adjustment factor. ; Directly call the pre-cached corresponding data in the microprocessor's high-speed registers The pulse duty cycle parameter is used to instantly update the PWM modulation parameters and reduce the operating power to a safe level. ,in Ensure the basic purification capability of the electric field; continuously monitor the high-voltage output voltage, and once the voltage recovers to the normal fluctuation range and does not trigger the flashover judgment threshold again, determine that the flashover discharge path has disappeared, and gradually increase the output power to the target operating power according to the safe power recovery slope. ;

[0157] F4, power restored to Then, reset the flashover status flag to 0; simultaneously, set the current pulse frequency, duty cycle, and output voltage. The synchronous feedback to the real-time execution process of electric field adaptive adjustment serves as the initial parameters for the next round of electric field adaptive frequency conversion pulse modulation, thus avoiding parameter conflicts.

[0158] During equipment operation, the microprocessor records the number of flashover triggers in real time, and outputs the flashover frequency data to the health diagnostic unit on an hourly basis for health assessment of the equipment's discharge characteristics.

[0159] VI. Diagnosis and Early Warning Mechanism Based on Health Vector

[0160] Based on the core operational data collected and calculated by the aforementioned units, this unit constructs a multi-dimensional health vector adapted to commercial kitchen fume purification equipment. It comprehensively characterizes the health level and degradation trend of the electric field throughout its entire life cycle. By setting preset industry-adaptability degradation thresholds, it enables early identification and graded warning of potential faults, transforming post-event maintenance into predictive maintenance, reducing the operation and maintenance costs of commercial kitchen equipment, ensuring long-term stable operation of the equipment, and matching the actual application needs of unattended and low-maintenance commercial kitchens.

[0161] The multidimensional health vector selects three health assessment dimensions for commercial kitchen fume purification equipment: the most critical, detectable in on-site operation and maintenance, and maintainable. The physical characteristics are easy to quantify, and the degradation threshold is determined through experimental calibration and on-site application verification in the fume purification industry, as shown in Table 1.

[0162] Table 1: Multidimensional Health Vector Assessment Dimensions, Features, Deterioration Judgment Thresholds, Corresponding Faults and Maintenance Measures

[0163] Evaluation Dimensions physical characteristic quantity Degradation threshold Corresponding fault type On-site operation and maintenance measures Impedance stability Impedance offset >15% Insulator carbonized heavy oil residue Clean the plates and replace the insulators Corona effect Attenuation factor <0.7 Needle passivation tip loss Grind the electrode needle and replace the corona electrode. Discharge characteristics flashover frequency >10 times / h Interval offset and foreign object attachment Correcting electrode spacing and removing foreign objects from the electric field

[0164] Based on the degree of equipment failure and degradation, a two-level early warning rule is set. The early warning output method matches the actual needs of on-site display and remote operation and maintenance in commercial kitchens, clearly defining the handling methods for different levels, and balancing the timeliness of operation and maintenance with the continuity of equipment operation.

[0165] Level 1 Warning (Maintenance Reminder): If a single-dimensional feature value reaches 80% of the degradation threshold, or if there is an abnormal load identification or sensor sampling failure, a Level 1 warning will be triggered directly. The Level 1 warning will only display the maintenance reminder on the equipment's human-machine interface and will not affect the normal operation of the equipment. If the feature value is below 70% of the degradation threshold for three consecutive control cycles, the microprocessor will automatically clear the maintenance reminder to avoid false warnings.

[0166] Level 2 Warning (Fault Alarm): If a single-dimensional or multi-dimensional feature exceeds the degradation threshold, or if oil pollution accumulates beyond expectations or flashover frequency exceeds 10 times for 1 consecutive hour, a Level 2 warning will be triggered directly. The Level 2 warning will be highlighted on the human-machine interface and a switch signal will be output to the remote operation and maintenance platform to prompt on-site operation and maintenance personnel to perform emergency maintenance to avoid the failure from worsening and causing a decrease in purification efficiency and equipment damage. After the Level 2 warning is triggered, the equipment must be physically maintained and manually reset through the human-machine interface. The alarm signal will then terminate simultaneously to ensure that maintenance is in place.

[0167] VII. Hardware System Implementation and Collaborative Operation Process

[0168] This embodiment also provides an intelligent control system for oil fume purification with adaptive electric field adjustment function. As the hardware implementation carrier of the aforementioned control method, all hardware modules are selected from industrial-grade components, which are fully matched to the on-site application environment of commercial kitchens with high temperature, dust, strong electromagnetic interference, and high humidity. It is deeply coordinated with the aforementioned control method in terms of software and hardware, and fully executes all functional logic to realize real-time adaptive adjustment of the electric field parameters for oil fume purification, ensuring stable, efficient, safe, and low-ozone operation of the equipment under complex working conditions in commercial kitchens.

[0169] 7.1 Composition and Functions of Core Hardware Modules

[0170] The core of this system comprises four main modules. The hardware configuration, functional boundaries, and corresponding relationships between each module and the aforementioned control methods are clearly defined, and all modules are adapted to the industry application scenarios of commercial kitchens, as detailed below:

[0171] Signal acquisition module: This is the "data sensing terminal" of the control method of this invention. It integrates a high-temperature resistant integrated temperature sensor, an anti-interference capacitive humidity sensor, a precision resistive voltage divider DC voltage sampling circuit, a non-contact high-frequency current transformer (HFCT), and a signal processing link. The hardware protection level is ≥IP54, making it suitable for dusty and high-humidity kitchen environments. It is used to accurately acquire basic parameters such as electric field intake temperature, kitchen ambient humidity, high-voltage power supply output voltage, and micro-current fluctuation signals. After impedance matching and noise filtering, the data is transmitted to the main control module, providing highly reliable raw data support for subsequent dynamic benchmark modeling, load characteristic identification, and other full-process calculations.

[0172] Main control module: This is the "core computing and control hub" of the system. It has a built-in STM32H743IIK6 industrial-grade microprocessor, and is equipped with non-volatile memory and high-speed registers. The non-volatile memory stores all the calibration parameters and execution programs required by the aforementioned control methods. It is responsible for performing the entire process of dynamic reference impedance modeling, load characteristic spatiotemporal decoupling identification, adaptive frequency conversion pulse modulation, flashover avoidance, health diagnosis and early warning according to the preset control cycle. It outputs PWM drive instructions and early warning control signals, realizes the timing coordination and priority scheduling of each module, and has strong anti-electromagnetic interference capability to avoid electromagnetic interference from the kitchen power grid and equipment.

[0173] High-voltage power supply module: The "execution terminal" that supplies power to the fume purification electric field. It uses a commercial-grade industrial high-voltage power supply specifically designed for fume purification and is electrically connected to the fume purification electric field. It receives PWM control commands from the main control module, precisely adjusts the pulse output frequency, duty cycle, and DC output voltage, and outputs high-voltage pulse signals to the electric field that meet the current kitchen operating conditions. It is compatible with the aforementioned differentiated control rules for all operating conditions and flashover avoidance energy regulation requirements. It has overcurrent, overvoltage, and overtemperature protection functions, realizing the synergy of adaptive voltage regulation and safe operation of the electric field, and matching the continuous operation requirements of commercial kitchens.

[0174] Human-Machine Interaction Early Warning Module: This module serves as the "status display and early warning output terminal" of the system. It utilizes a waterproof and dustproof industrial-grade touchscreen and is strongly connected to the main control module via RS485 wired communication. It displays real-time parameters such as equipment operating status, absolute value of kitchen fume impedance offset rate, and estimated ozone generation. It simultaneously outputs the aforementioned two levels of early warning information and supports functions such as fault reset, operating parameter query, and historical data export, meeting the needs of on-site maintenance and remote monitoring. It also reserves a 485 / Modbus communication interface for connection to a commercial kitchen smart maintenance platform, enabling remote monitoring and management of the equipment.

[0175] 7.2 System Collaborative Operation Process

[0176] After the system is powered on, each module operates collaboratively according to a preset timing sequence:

[0177] S1. The signal acquisition module immediately enters the real-time monitoring state, completes the accuracy calibration and initialization of all sensors and sampling circuits, and ensures the accuracy of the raw data acquisition.

[0178] S2. The main control module starts the full-process operation according to the preset control cycle, and sequentially completes the dynamic reference impedance modeling, load characteristic spatiotemporal decoupling identification, and adaptive frequency conversion pulse modulation parameter calculation.

[0179] S3. The main control module outputs PWM drive commands to the high-voltage power supply module, driving the high-voltage power supply module to perform adaptive electric field adjustment and apply high-voltage pulses that conform to the current kitchen working conditions to the oil fume purification electric field.

[0180] S4. The main control module monitors the flashover status in real time through hardware interrupts. If a flashover discharge is triggered, it immediately performs a flashover avoidance action to suppress the flashover discharge and ensure the continuous operation of the equipment.

[0181] S5. The main control module performs real-time equipment health diagnosis based on the full-process operation data. If the characteristic quantity reaches the warning threshold, the corresponding level of warning is triggered, which is displayed in the human-machine interaction warning module and pushed to the remote operation and maintenance platform.

[0182] S6. All operating parameters, early warning information, and fault records are stored in real time in the non-volatile memory of the main control module. They can be queried or exported through the human-machine interaction early warning module, providing data support for on-site operation and maintenance.

[0183] The above description is merely a preferred embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the principles and spirit of the present invention should be included within the scope of protection of the present invention.

[0184] Furthermore, the various embodiments of the present invention are not independent of each other. Those skilled in the art can combine the technical features of the various embodiments according to actual needs, and the combined technical solution still falls within the protection scope of the present invention.

Claims

1. A method for controlling and purifying oil fume with adaptive electric field adjustment function, characterized in that, The method includes the following steps: Real-time acquisition of inlet air temperature, ambient humidity, high voltage output voltage and current micro-fluctuation signals of the oil fume purification electric field; A dynamic reference impedance model is constructed that integrates logarithmic aging decay of equipment, temperature and humidity environment correction, and impedance offset of normal oil pollution on the electrode plate to obtain the real-time dynamic reference impedance of the electric field. Based on the dynamic reference impedance and current micro-fluctuation signal, the impedance offset rate and current fluctuation entropy are calculated, and a two-dimensional feature space is constructed with the impedance offset rate as the horizontal axis and the current fluctuation entropy as the vertical axis for hierarchical judgment to distinguish between high concentration of oil fume load, high humidity interference, clean low load and abnormal state of unexpected oil accumulation. Based on the load determination results and the absolute value of the impedance offset rate, with the optimization goal of suppressing ozone generation, under the condition of meeting the minimum purification power constraint and hardware safety boundary, the pulse frequency, duty cycle and DC output voltage of the high voltage power supply are adaptively adjusted according to the priority of increasing pulse frequency, decreasing pulse duty cycle and then decreasing DC output voltage. The voltage drop rate of the high voltage output is monitored in real time. When the drop rate exceeds the preset flashover judgment threshold, the corresponding energy reduction ratio factor is called based on the absolute value of the current impedance deviation rate. The target working power is reduced to the safe power while maintaining the basic purification power of the electric field. After the voltage stabilizes, it is restored according to the safe slope. By combining impedance offset rate, corona performance attenuation factor, and flashover frequency, a multidimensional health vector is constructed to achieve graded diagnosis and predictive maintenance of equipment status.

2. The oil fume purification and control method with adaptive electric field adjustment function as described in claim 1, characterized in that, The expression for the dynamic reference impedance model is: ,in, Static impedance Based on the cumulative operating time of the equipment The logarithmic aging decay function, The aging factor is... This is the time scaling factor. Based on intake air temperature Ambient humidity Temperature and humidity correction operator, Based on historical cumulative oil fume load The impedance offset of the electrode plate under normal oil contamination was obtained through online learning.

3. A method for controlling and purifying oil fume with adaptive electric field adjustment function as described in claim 1 or 2, characterized in that, The impedance offset The calculation formula is: , The real-time measurement impedance of the electric field for oil fume purification; The calculation process of the current fluctuation entropy is as follows: A sequence of micro-current fluctuation signals is extracted using a sliding window of predetermined duration. The least squares method is used to fit the sequence and remove the linear trend term to obtain a quasi-steady-state current fluctuation sequence. The quasi-steady-state current fluctuation sequence is then divided into equal parts, and the distribution probability of data points in each interval is calculated. Finally, the information entropy formula is applied... The entropy of current fluctuation is obtained. , This represents the number of steady-state fluctuation intervals of a single-segment quasi-steady-state current. For the first The probability distribution of the interval.

4. The oil fume purification and control method with adaptive electric field adjustment function as described in claim 1, characterized in that, The hierarchical determination in the two-dimensional feature space is based on the critical value of impedance background fluctuation. Current fluctuation entropy discrimination threshold Execution, the specific judgment rules are as follows: like >0 indicates an abnormal state of unexpected accumulation of oil stains on the electrode plates; like <0 and > and > It was determined to be a high-concentration oil fume load; like < and > and ≤ High humidity environment interference; like ≤ and The current fluctuation range is within the baseline fluctuation range of the clean low-load operating condition as specified in Experiment 2, and is judged to be a clean low-load / normal state.

5. The oil fume purification and control method with adaptive electric field adjustment function as described in claim 1, characterized in that, The method also includes a linear prediction function for ozone generation: In the formula, This is the ozone generation ratio coefficient; This is the output voltage of the high-voltage power supply; This is the critical corona initiation voltage of the electric field used for fume purification. This refers to the pulse duty cycle. The pulse frequency; ozone generation and excess corona induction voltage. and Positive correlation Negative correlation.

6. The oil fume purification and control method with adaptive electric field adjustment function as described in claim 2, characterized in that, The minimum purification power constraint is implemented based on the minimum purification power constraint model, the model expression of which is: , This refers to the rated no-load power under clean operating conditions. The power gain coefficient of the oil fume load is determined based on the Deutsch formula; This represents the absolute value of the impedance offset rate. Electric field average output power model: , It is the power conversion constant; The parameter adjustment process must meet the following requirements: And all initial baseline parameters are based on Complete aging correction.

7. The oil fume purification and control method with adaptive electric field adjustment function as described in claim 1, characterized in that, The energy down-regulation factor The absolute value of the impedance offset rate is It has a piecewise linear mapping relationship: When the value is between 0.1 and 0.3, Take 0.6; When it is between 0.3 and 0.5, Take 0.7; When it is between 0.5 and 0.8, Take 0.8; The safe power , For the target operating power, and .

8. The oil fume purification and control method with adaptive electric field adjustment function as described in claim 7, characterized in that, The step of reducing the target operating power to a safe power is achieved by calling the corresponding pre-cached register in the microprocessor's high-speed register. The duty cycle parameter enables microsecond-level response; The term "waiting for voltage stabilization" specifically refers to the high-voltage output voltage returning to its normal fluctuation range without flashover triggering, and the safe power recovery slope during power recovery being less than 1% of the maximum power recovery slope.

9. The oil fume purification and control method with adaptive electric field adjustment function as described in claim 2, characterized in that, The impedance offset of the electrode plate under normal oil contamination Learn through online training; after the initial operation of the equipment or physical cleaning of the plates, and historical cumulative oil fume load Set to zero; within the maintenance time window when the equipment has no oil fume load, calculate the difference between the real-time measured impedance of the electric field and the theoretical intrinsic impedance under clean conditions, as the impedance offset introduced by oil accumulation, and store the impedance offset data in non-volatile memory to form an empirical dataset.

10. The system of the oil fume purification and control method with adaptive electric field adjustment function as described in any one of claims 1, characterized in that, The system comprises a signal acquisition module, a main control module, a high-voltage power supply module, and a human-machine interaction early warning module working together. The signal acquisition module integrates a temperature sensor, a humidity sensor, a precision resistor voltage divider DC voltage sampling circuit, and a non-contact high-frequency current transformer. It is used to acquire in real time the intake air temperature, ambient humidity, high voltage output voltage, and current micro-fluctuation signals, and to complete signal conditioning and calibration. The main control module is an industrial-grade microprocessor with built-in non-volatile memory and high-speed registers. It stores all calibration parameters and execution programs for dynamic reference impedance modeling, load characteristic determination, ozone suppression-oriented parameter adjustment, flashover avoidance, and multi-dimensional health vector diagnosis. It is responsible for the entire process calculation and generates control commands. The high-voltage power supply module is a controllable industrial-grade high-voltage power supply that receives control commands from the main control module, precisely adjusts the pulse frequency, duty cycle, and DC output voltage, and has hardware protection against overcurrent, overvoltage, and overtemperature. The human-machine interaction early warning module is a waterproof and dustproof industrial-grade touch screen that communicates with the main control module to realize the display of operating parameters, output of hierarchical early warning information, data query and remote communication. The modules work together to complete the adaptive adjustment of the electric field parameters of the fume purification and the full life cycle management of the equipment.

Images