Dynamic impedance self-adaptive system for food lossless oiling

Abstract

The invention relates to the technical field of food industrial processing and intelligent manufacturing equipment, and discloses a dynamic impedance self-adaptive system for food lossless oiling, which comprises a controller and a plurality of conductive electrodes arranged around an atomized oil path, on one hand, the oil injection actuator is adjusted in a closed-loop mode based on the sum of the signals so as to stabilize the mass flow rate, on the other hand, the spatial distribution symmetry of spraying is diagnosed in parallel based on the difference between the signals, an information homologous mechanism of process feedback and diagnosis is established, and the system can immunize the oil product viscosity in principle; and due to process disturbance caused by the change of the feeding pressure and the state of the nozzle, the on-line self-diagnosis capability of the system on the execution quality of the spraying process is improved, so that the accuracy and the stability of the process are ensured.

Description

Technical Field

[0001] This invention relates to a dynamic impedance adaptive system for non-destructive oiling of food, belonging to the technical field of food industrial processing and intelligent manufacturing equipment. Background Technology

[0002] Currently, using programmable logic controllers (PLCs) or embedded systems to send precise electrical signal commands to various actuators to achieve automated control of material flow is a widely adopted and technologically mature basic operating method in the industry. This method decomposes complex processes into a series of deterministic control commands and execution actions, making large-scale standardized continuous production possible. However, when this command-driven control method is applied to the specific process of fluid spraying, especially in production scenarios that pursue long-term, high stability, and high quality, an inherent limitation begins to emerge. In a real production environment, the actual mass flow rate output of the spraying actuator, as a fluid machine, is continuously affected by a series of dynamically changing and unpredictable physical factors. For example, subtle changes in workshop temperature cause fluctuations in oil viscosity, long-term operation leads to the formation of tiny oil deposits inside the nozzle, causing partial blockage, and the compressed air pipeline network experiences instantaneous pressure fluctuations due to the start and stop of other equipment. These factors together constitute a complex dynamic process impedance, resulting in a dynamic and unpredictable nonlinear deviation between the electrical signal commands issued by the controller and the actual amount of oil sprayed from the nozzle.

[0003] To address the aforementioned challenges, those skilled in the art have attempted several improvement approaches, such as constructing a multivariable compensation control model by adding online viscometers, mass flow meters, and pressure sensors. However, this approach not only increases the hardware cost and maintenance complexity of the system, but the response speed and robustness of its multi-sensor data fusion algorithm itself pose new technical challenges. This disconnect between control commands and physical execution results is not only evident in complex automated spraying systems, but also in seemingly simple oiling tools, where inherent control logic defects are similar, or even more fundamental. For example, in the case of authorization announcement number C... Chinese invention patent N204363270U discloses a novel food coating brush that integrates an oil reservoir into the brush body and sets permeation holes and scales to achieve so-called quantitative coating. However, this design is essentially a completely open-loop passive feeding system. Its actual oil output rate is entirely subject to the gravity and viscosity of the oil itself, as well as the operator's actions. It cannot form any closed-loop feedback, cannot actively respond to dynamic disturbances in the production environment, and cannot monitor and adjust the uniformity of coating online. This is completely unacceptable in industrial continuous production scenarios that pursue high precision and high consistency.

[0004] Specifically, existing technologies suffer from the following shortcomings: 1. There is a lack of a feedback loop between the control system's commands and the physical world's execution results, which can directly, in real time, and comprehensively reflect the impact of all process disturbances. The control system is essentially blind to the true working state of its actuators. 2. To avoid product quality problems caused by execution distortion, production practices commonly employ compensatory strategies such as increasing the average spraying amount, increasing production line inspection frequency, or periodically replacing nozzles. This results in long-term hidden waste of raw materials and excessive reliance on manual experience. Therefore, the technical problem this invention aims to solve is how to construct a system that can penetrate the fog of process impedance in real time, enabling the oil injection actuator to possess closed-loop adaptive capability to execution deviations, thereby ensuring that control commands can be accurately reproduced by the physical world without increasing system complexity. Summary of the Invention

[0005] This invention provides a dynamic impedance adaptive system for non-destructive oiling of food. Its main purpose is to solve the problem that the existing control method has a lack of direct feedback on the dynamic process impedance, which leads to a deviation between the control command and the actual oil injection quantity.

[0006] To achieve the above objectives, the present invention provides a dynamic impedance adaptive system for non-destructive oil coating of food, comprising an oil spraying actuator, an electrical sensing device, and a controller. The controller controls the oil spraying actuator via a drive signal. The electrical sensing device includes multiple electrically isolated conductive electrodes arranged around the path of the atomized oil sprayed from the oil spraying actuator. The controller is configured to execute the following rules: Rule a: Multiple real-time sensing signals are acquired through multiple conductive electrodes, wherein each real-time sensing signal is generated by the space charge carried by the atomized oil due to triboelectric charging during the atomization process and induced on the corresponding conductive electrode. Rule b involves continuously comparing the sum of multiple real-time sensor signals with a preset target value representing the target mass flow rate, and adjusting the drive signal to the fuel injection actuator in real time based on the deviation generated by the comparison using a preset control algorithm. Rule c generates a diagnostic signal characterizing the spatial distribution symmetry of atomized oil based on the numerical differences between multiple real-time sensor signals, and uses the diagnostic signal to trigger a preset alarm or control command.

[0007] Preferably, in rule b, the controller is configured to: integrate or average multiple real-time sensor signals in each control cycle to obtain a sum; and use a proportional-integral control algorithm or a proportional-integral-derivative control algorithm to calculate and update the drive signal for the fuel injector actuator based on the deviation between the sum and a preset target value, wherein the drive signal is a voltage signal or pulse width modulation signal used to control the internal fluid valve of the fuel injector actuator.

[0008] Preferably, the controller is further configured to: perform long-cycle statistical trend analysis on the drive signal output to the fuel injection actuator in rule b; the statistical trend analysis includes calculating the moving average of the drive signal on a preset slow time scale longer than the control cycle; and generating a predictive maintenance instruction related to the deterioration of the fuel injection actuator's health status when the moving average deviates continuously in one direction relative to a baseline value characterized by the health status of the fuel injection actuator through learning operation or a preset method.

[0009] Preferably, in rule c, the controller is configured to: select at least two first and second conductive electrodes, respectively located on opposite sides of the atomized oil path, from a plurality of conductive electrodes; and perform calculations. To calculate the symmetry index, where, The symmetry index The amplitude of the real-time sensing signal generated by the first conductive electrode. The amplitude of the real-time sensing signal generated by the second conductive electrode is used; and a diagnostic signal is generated when the absolute value of the symmetry index continuously exceeds a preset symmetry instability threshold.

[0010] Preferably, the system also includes a controllable high-voltage power supply connected to multiple conductive electrodes; and the controller is configured to operate in a time-division multiplexing mode: within one drive cycle, the controllable high-voltage power supply is controlled to apply a preset voltage to multiple conductive electrodes, thereby actively changing the spatial distribution of the atomized oil through the resulting electrostatic field; and in the immediately following sensing cycle, the controllable high-voltage power supply is disconnected, and rules a, b and c are executed.

[0011] Preferably, the system further includes an active detection device located upstream of the fuel injection actuator. The active detection device includes an ion generator for generating a stable ion wind and a sensing electrode for collecting ions that penetrate the food or are not captured by the food. The controller is also configured to generate a feedforward signal characterizing the wettability of the food surface based on the signal from the sensing electrode, and to dynamically adjust a preset target value using the feedforward signal before performing the comparison of rule b.

[0012] Preferably, the multiple conductive electrodes of the electrical sensing device are four electrically isolated sector-shaped conductive electrodes, which together form a ring structure and are coaxially located in the path of the atomized oil.

[0013] Preferably, the system further includes multiple signal amplification circuits, which are respectively connected to multiple conductive electrodes. Each signal amplification circuit is a current-to-voltage amplification circuit with high input impedance, configured to convert the induced current on the corresponding conductive electrode into a real-time sensing signal in voltage form.

[0014] Preferably, the preset alarm or control command triggered by the diagnostic signal in rule c is specifically: an alarm signal indicating that the fuel injector actuator has experienced partial blockage, or a safety command to stop the fuel injector actuator from operating.

[0015] Preferably, the controller is further configured to: perform a fast Fourier transform on the raw waveform data of the real-time sensing signal generated by any conductive electrode during the sensing period to obtain its spectrum; and generate a medium state signal based on the ratio of harmonic energy to fundamental frequency energy in the spectrum, and use the medium state signal to trigger an early warning indicating the deterioration of the chemical properties of the oil.

[0016] Compared with the prior art, the beneficial effects of the present invention are: 1. This invention establishes a direct correlation between the triboelectric charge generated during the atomization process and the actual mass flow rate. By continuously correcting the deviation between the charge sensing signal and the target value through the controller, the adjustment of the system drive signal is no longer an indirect control based on a preset model, but a real-time response to the comprehensive influence of all process disturbances, including oil viscosity, feed pressure and nozzle status, so that the mass flow rate output of the fuel injector remains stable in a dynamically changing working environment.

[0017] 2. This invention divides a single sensing loop into multiple isolated conductive electrodes and combines them with the parallel analysis of the differences in the signals of each electrode by the controller. This enables the system to monitor the spatial distribution symmetry of atomized oil while controlling the total mass flow rate in a closed loop. When a local blockage occurs in the nozzle, causing abnormal spray pattern but no change in total flow, the imbalance of the signals of each electrode provides a basis for judging the quality of process execution, avoiding potential product quality hazards that may occur when only the total amount is used as the control target.

[0018] 3. In addition to performing real-time adjustments to the drive signal of the fuel injector actuator, the present invention utilizes the controller itself to perform long-term statistical trend analysis on the output drive signal. In the process of a healthy fuel injector gradually deteriorating and requiring the controller to continuously increase the drive output to maintain a constant fuel injection quantity, the long-term mean drift of the drive signal becomes an intrinsic indicator of its health status. This method utilizes the secondary information that is inevitably generated during the operation of the control system, and provides a basis for predictive maintenance decisions for the system without adding any physical sensors. Attached Figure Description

[0019] Figure 1 This is a functional block diagram of the closed-loop control and state diagnosis logic of the system of the present invention; Figure 2 This is a system response data curve diagram under the condition of partial nozzle blockage according to the present invention; Figure 3This is an interactive sequence diagram of the baseline self-learning process of the controller parameters in this invention. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be described in detail below. Obviously, the described embodiments are only some embodiments of this invention, and not all embodiments. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0021] The present invention discloses a dynamic impedance adaptive system for non-destructive oil coating of food, which mainly consists of an oil spraying actuator, an electrical sensing device, and a controller. The electrical sensing device is installed downstream of the oil spraying actuator, while the controller establishes electrical signal connections with both the electrical sensing device and the oil spraying actuator to form a complete closed-loop feedback control circuit. In industrial environments, automated oil spraying processes often face unpredictable process disturbances such as the influence of ambient temperature on oil viscosity, changes in supply pressure due to pipeline load variations, and minor nozzle blockages due to long-term operation. These disturbances affect the electrical signal commands issued by the control system. There is a dynamic deviation between the actual amount of oil sprayed from the nozzle and the actual amount of oil sprayed. This solution obtains the charge carried by the atomized oil in space due to the triboelectric effect and uses it as a feedback signal that can reflect the influence of all process disturbances in real time. Based on the deviation between this feedback signal and the preset target value, the controller makes millisecond-level real-time adjustments to the drive signal of the oil injection actuator. At the same time, the controller also performs online diagnosis of the symmetry of the spray pattern based on the difference in the spatial distribution of the feedback signal. Thus, while ensuring the stability of the spray quality flow rate, the system provides an online self-diagnostic capability for its own execution quality.

[0022] The electrical sensing device specifically consists of four electrically isolated sector-shaped conductive electrodes. These four electrodes physically form a hollow ring structure, isolated from the electrical grounding point of other equipment on the production line. They are coaxially arranged along the path of the atomized oil sprayed by the fuel injector. Each individual sector-shaped conductive electrode is connected via a coaxial cable to an independent high-input-impedance current-to-voltage amplifier circuit, which also serves as a signal amplification circuit. It should be noted that when the fuel injector, such as a pneumatic proportional valve, atomizes the oil into micron-sized droplets under high-pressure gas, the droplets experience intense friction with the nozzle inner wall, the nozzle opening, and the air. Based on the physical principle of triboelectric charging, the oil mist carries... Carrying a net charge, a space charge cloud is formed that is statistically highly positively correlated with the number of ejected oil droplets, i.e., the mass flow rate. When this charged oil mist cloud passes through a ring structure composed of four fan-shaped conductive electrodes, an equal amount of opposite charge is induced on the surface of each conductive electrode, forming an induced current in the picoampere to nanoampere range. This induced current is converted into a real-time sensing signal in the form of a voltage with an amplitude in the millivolt or volt range, which is convenient for subsequent processing, by a corresponding high input impedance current-to-voltage amplifier circuit. In this way, the system utilizes the physical phenomena that accompany the oil atomization process to transform a fluid dynamics problem affected by multiple variables into a control problem that can be monitored by analyzing four independent electrical signals.

[0023] The controller, which can be a standard industrial-grade programmable logic controller (PLC) or an embedded microcontroller system, is configured to execute the closed-loop feedback control defined by rule b and the symmetry diagnostics defined by rule c in parallel. When executing rule b, within a control cycle of no more than 10 milliseconds, the controller synchronously acquires all four real-time sensor signals from the outputs of four signal amplifier circuits through its multi-channel analog input interface, and sums the amplitudes of these four signals to obtain a signal sum that characterizes the current total mass flow rate. For example, within a specific control cycle, the voltage signal amplitudes of the four channels acquired by the controller are respectively... Then the sum of the calculated signals is The controller internally runs a proportional-integral (PI) control algorithm and uses the sum of this signal as the algorithm's process variable, along with a preset target value, such as a value also calibrated in voltage units representing the target mass flow rate. The PI control algorithm performs continuous comparisons based on the deviation between the two values. ,Right now This is used to calculate and update a drive signal for the fuel injection actuator. Specifically, this drive signal can be a 0-10V DC voltage signal that controls the opening of the fluid proportional valve inside the fuel injection actuator, or a pulse width modulation (PWM) signal with a fixed period and adjustable duty cycle. When the deviation... When the value is positive, it indicates that the actual injection quantity is lower than the target value. The PI controller will correspondingly increase the duty cycle of its output voltage signal or PWM signal to increase the valve opening. Conversely, it will decrease the valve opening. Through this millisecond-level continuous negative feedback regulation, the controller can actively counteract the total flow disturbance caused by changes in oil viscosity, fluctuations in feed pressure, or uniform wear of the nozzle, and stably lock the actual mass flow rate at the target value.

[0024] Simultaneously, the controller executes rule c in parallel, which involves diagnosing the spatial distribution symmetry of the atomized e-liquid using four independent real-time sensor signals. Specifically, the controller selects a pair of electrodes located on opposite sides of the atomized e-liquid path from the four fan-shaped conductive electrodes. For example, the electrode on the upper side in the vertical direction is defined as the first conductive electrode, and the amplitude of its generated real-time sensor signal is [value missing]. The lower electrode is the second conductive electrode, and the amplitude of the real-time sensing signal it generates is... Subsequently, the controller performs calculations. To calculate a dimensionless symmetry index, where, The symmetry index The amplitude of the real-time sensing signal generated by the first conductive electrode. The amplitude of the real-time sensing signal generated by the second conductive electrode is used to reduce the impact of instantaneous signal fluctuations. and The value is the average of multiple sampling points within a 100-millisecond time window; under a symmetrical spray pattern, and The magnitudes should be very close, so that the symmetry index... The absolute value of fluctuates within a very small range close to zero, for example... The baseline value of this fluctuation range can be determined by performing a baseline learning operation of standard spraying after initial system installation or replacement of a new nozzle; the controller has a preset symmetry instability threshold, for example, 0.1, which is triggered when the controller detects a symmetry index... If the absolute value of the value continuously exceeds the threshold, for example, if it is greater than 0.1 for 5 consecutive calculation cycles, the controller will determine that the oil injection actuator has experienced local blockage that causes asymmetrical spray pattern, and immediately generate a diagnostic signal. This diagnostic signal can be used to trigger an alarm signal on a human-machine interface, or directly trigger a safety command to stop the oil injection actuator from working, thereby avoiding mass product quality problems caused by uneven spraying.

[0025] To further enhance system reliability, the controller is also configured to perform long-cycle statistical trend analysis on the drive signals output to the fuel injection actuators in rule b, enabling predictive maintenance of the actuators' health status. This function utilizes the controller's existing computing resources by adding a purely software-based health status observer module within the controller. This module calculates the moving average of the drive signals on a slow timescale, such as every minute, much longer than the control cycle. On a newly installed fuel injection actuator, the controller performs a baseline learning operation, recording the moving average of the drive signals required to maintain the target mass flow rate under standard operating conditions. The average value is stored as the baseline value for the health status, such as 5.2V. During subsequent long-term operation, as the nozzles slowly and progressively become clogged, the PI controller will inevitably and continuously increase the average value of its output drive signal in order to maintain a constant injection volume. When the health status observer module detects that the moving average value of the drive signal has a continuous unidirectional deviation from the baseline value, for example, when the moving average value for one consecutive hour exceeds 1.2 times the baseline value, i.e., 6.24V, the controller will generate a predictive maintenance instruction related to the deterioration of the health status of the injection actuator, providing a basis for decision-making for planned maintenance before equipment failure.

[0026] In applications requiring high production flexibility, to achieve precise spraying of food of different sizes, the system may also include a controllable high-voltage power supply connected to multiple conductive electrodes. In this case, the controller is configured to operate in time-division multiplexing mode. Within one drive cycle, for example, a 9-millisecond time window, the controller disconnects each conductive electrode from the signal amplification circuit and instead controls the controllable high-voltage power supply to apply a uniform DC high voltage ranging from 0 to 2000V to all conductive electrodes. Since the atomized oil droplets and conductive electrodes typically carry the same charge, this voltage creates a repulsive electrostatic field in the central region of the loop, generating radial compressive force on the oil mist, thereby effectively narrowing the spray cone. The higher the voltage... The higher the focusing effect, the narrower the spray coverage. Through this electronic method, the system can actively and quickly adjust the spray pattern. In the following sensing cycle, for example, a time window of 1 millisecond, the controller disconnects the controllable high-voltage power supply and quickly reconnects each conductive electrode to its corresponding signal amplification circuit. During this brief field-free window, the aforementioned rules a, b, and c are executed to complete a measurement and update of the actual charge flow. By switching between the driving and sensing modes at a frequency of, for example, 100Hz, the system achieves control of the fuel injection quantity and dynamic shaping of the spray pattern without adding mechanical adjustment components.

[0027] When the system operates in time-division multiplexing mode, to reliably isolate the high voltage applied during the drive cycle from the picoampere-level current acquired during the sensing cycle, a high-speed solid-state analog switch directly driven by the controller is connected in series between each conductive electrode and the corresponding high-input-impedance current-to-voltage amplifier circuit. An overvoltage protection circuit consisting of a series current-limiting resistor and a parallel transient voltage suppression diode array is set at the signal input terminal of each amplifier circuit. When the controller performs a switching operation, its internal logic is set to enforce a 50 to 100 microsecond silent delay after the command to disconnect the controllable high-voltage power supply is issued and before the command to close the solid-state analog switch to turn on the signal amplifier circuit is issued. This delay time is used to discharge the residual charge on the conductive electrode, so that the input terminal of the amplifier has returned to a stable baseline state at the beginning of the sensing cycle.

[0028] To incorporate the final coating quality into the control, the system may also include an active detection device located upstream of the fuel injection actuator. This active detection device includes a corona discharge needle for generating a stable ion wind and a metal sensing electrode opposite it, mounted below the conveyor belt. When no food passes through, the sensing electrode receives a stable baseline ion current. When a biscuit with a drier, more porous surface passes by, its surface adsorbs and neutralizes a large number of ions, resulting in an ionic current that eventually reaches the sensing electrode. As the current decreases, the controller calculates the rate of current decay in real time. This generates a feedforward signal that can quantify the wettability of food surfaces in real time and without contact. Before executing the comparison in rule b, the controller uses this feedforward signal to dynamically adjust a preset target value. For example, the dynamic target value can be adjusted to... ,in, This is the preset target value after dynamic adjustment. As the baseline target value, It is a gain coefficient. This is a real-time infiltrative index. It is the historical average wetting index. Thus, when a cookie with strong oleophilicity is about to enter the spraying area, the system will proactively reduce the target amount of oil sprayed this time, thereby achieving adaptive coating for different individual foods.

[0029] Finally, to achieve online monitoring of the chemical state of the oil itself as a process medium, the controller can also be configured to perform a Fast Fourier Transform (FFT) on the raw waveform data of the real-time sensing signal generated by any conductive electrode within the sensing cycle of the time-division multiplexing mode to obtain its spectrum. For a chemically stable oil, the generated sensing pulse waveform is statistically stable, with its energy mainly concentrated in the fundamental frequency band. However, for an oil that has deteriorated due to oxidation and polymerization over a long period of use, the increased charge-to-mass ratio and spatial non-uniformity will result in more high-frequency noise and harmonic components in the sensing pulse waveform. The controller generates a medium state signal that directly characterizes the degree of oil deterioration by calculating a harmonic energy ratio (HER), i.e., the ratio of high-frequency energy to fundamental frequency energy. When this HER value continuously increases relative to the baseline value of new oil and exceeds a preset threshold... When the system detects the deterioration of the oil's chemical properties, it triggers an early warning, indicating the need to replace the oil with new one. This extends the perception dimension of process control from the process execution itself to the quality monitoring of process materials. The frequency range of the fundamental and harmonic frequency bands used to characterize the deterioration of the oil's chemical properties is determined through a one-time calibration procedure performed during the system's initial operation. The specific steps of this procedure are as follows: First, with a brand-new nozzle installed on the injector and using undeteriorated standard oil, the system is run continuously for one minute under baseline conditions. During this period, the controller collects raw waveform data at a sampling rate of no less than 1 kHz and performs a fast Fourier transform on it to obtain a set of baseline spectrum data. Subsequently, the controller calculates and analyzes the energy distribution of this baseline spectrum data, defining the frequency point where the energy accumulates from zero Hz to 95% of the total energy as the upper limit frequency value of the fundamental frequency band. and will be higher Furthermore, the range below the Nyquist frequency is defined as the harmonic frequency band, and ultimately, the controller will use this... The values ​​are stored in a non-volatile memory module and used as a fixed basis for subsequent calculations of harmonic energy ratios.

[0030] Example 1: This example is a specific operational instance of the described technical solution in a particular industrial application scenario. In a production line operating continuously for 24 hours for automated oiling of high-end soda crackers, when the shift reaches the 16th hour, a nozzle of an oil spraying actuator, due to prolonged exposure to high temperatures while atomizing vegetable oil, begins to accumulate tiny carbonized oil deposits on one side edge of its nozzle. These oil deposits reduce the total flow cross-sectional area of ​​the nozzle, causing the actual mass flow rate of the atomized oil to fall below the target value set by the process. The controller collects the induced signals from four fan-shaped conductive electrodes and calculates the sum of these signals. The process variable was monitored relative to the preset target value. A negative deviation was generated; subsequently, the proportional-integral (PI) control algorithm inside the controller automatically increased the drive signal voltage to the fuel injector on a millisecond timescale based on this deviation, until the signal summed... Re-stabilized at the preset target value The system compensates for the decrease in total fuel injection volume caused by nozzle clogging. However, the location of the oil deposit is not at the center of the nozzle, which causes a distortion in the spatial distribution of the spray cone from symmetry to asymmetry. Specifically, more oil mist is sprayed onto the side where no oil deposit has formed. While the controller performs the aforementioned closed-loop control of the total mass flow rate, its internal parallel symmetry diagnostic logic continuously monitors the real-time sensing signal amplitude from the two conductive electrodes on opposite sides. and In comparison, as the asymmetry of the spray pattern intensifies, the symmetry index calculated by the controller increases. The absolute value, starting from a... The baseline state fluctuates within the range, gradually climbs and eventually exceeds the preset symmetry instability threshold of 0.1. In this scenario, the total amount feedback control of rule b avoids defects caused by insufficient oiling by compensating for the total flow, providing time for subsequent processing, while the symmetry diagnosis of rule c identifies information about the deterioration of process execution quality that the total amount control logic itself cannot perceive.

[0031] When the symmetry index After the absolute value of the spray pattern is greater than 0.1 for five consecutive calculation cycles, the controller determines that the spatial distribution symmetry of the spray has deviated significantly and generates a diagnostic signal. This diagnostic signal triggers two preset control commands: first, the human-machine interface of the central monitoring station on the production line displays an alarm signal indicating that the nozzle of spraying unit 3 is suspected of being partially blocked and that the spray pattern is asymmetrical; second, the yellow status indicator light above the spraying station is illuminated. Since the total spray volume is still maintained within the process requirements by the PI control algorithm, the system does not trigger an immediate production line shutdown, but instead allows the operator to remain in operation for a predetermined period. During material replenishment intervals, online inspections and nozzle replacements are performed on the fuel injection actuator. This approach transforms a progressive equipment failure that could potentially lead to product quality issues into a planned preventative maintenance operation, minimizing the impact on production cycle time. By processing single sensor information in parallel using both total summation and spatial difference dimensions, the system gains the ability to monitor the spatial symmetry of atomized fuel distribution while simultaneously controlling the total mass flow rate. Even under conditions where the actuator's health deteriorates, the system can maintain stable final process quality through the inherent coordination of control logic.

[0032] Example 2: To objectively verify the performance of the technical solution of this invention in dealing with dynamic process impedance commonly encountered in industrial production, this example was conducted. The purpose of the experiment was to quantitatively compare the ability of the system using the technical solution of this invention and the system using a traditional open-loop control method to maintain the stability of the spraying mass flow rate when faced with three typical process disturbances: changes in oil viscosity, fluctuations in feed pressure, and partial clogging of the nozzle. The experiment was conducted on a test platform simulating an industrial production environment. This platform mainly consists of an adjustable feed pressure and oil temperature oil supply system, a dynamic impedance adaptive system, and an electronic balance for measuring the actual spraying mass flow rate. The electronic balance has a measurement accuracy of 0.01g and a data refresh rate of 10Hz. The experiment was conducted in two groups: a control group and the present invention group. The control group used an open-loop control method, where the controller outputs a constant drive signal to the oil injection actuator, the value of which is the calibrated value of the target flow rate achievable under standard operating conditions. The present invention group, on the other hand, uses the complete functions of closed-loop feedback control and symmetry diagnosis based on real-time sensor signals. The target mass flow rate in the experiment was set to 50.0g / min, a parameter determined based on the coating amount per unit area requirements of typical food coating processes. The experiment simulated four operating conditions in sequence, starting with the oil temperature at 2... The system operates under the baseline condition of a feed pressure of 0.4 MPa. Subsequently, while keeping other conditions unchanged, the oil temperature is reduced to 15°C. To simulate the increase in oil viscosity caused by a drop in ambient temperature, the oil temperature was then restored to 25°C. The feed pressure was reduced to 0.3 MPa to simulate pressure fluctuations caused by changes in pipeline load. Finally, based on the restoration of the baseline operating conditions, a small obstacle was artificially set on one side of the nozzle to simulate a local blockage caused by oil deposits. Under each operating condition, the system was run continuously for 5 minutes, and the average value of the actual mass flow rate and symmetry index during the steady-state phase was recorded. The specific test data are recorded in Table 1 below.

[0033] Table 1: Performance comparison test data under different working conditions.

[0034] Under the baseline conditions of test number 1, both sample groups achieved a target mass flow rate close to 50.0 g / min, and the symmetry index of the sample group of this invention was [not specified]. Approaching zero; when entering test sequence 2, the increased oil viscosity led to increased flow resistance, the actual injection volume of the control group decreased by 15.2%, while the controller of the present invention's sample group automatically increased the drive signal output through the PI control algorithm, maintaining the actual injection volume at 49.8 g / min, with a deviation from the target value of less than 0.5%; similarly, under the condition of decreased feed pressure in test sequence 3, the injection volume of the control group decreased by 9.8%, while the injection volume of the present invention's sample group was basically unaffected; under the condition of partial nozzle blockage in test sequence 4, the injection volume of the control group decreased slightly, while the total quantity control loop of the present invention's sample group still maintained the total mass flow rate at the level of 50.1 g / min through compensation, and its symmetry diagnostic loop monitored the symmetry index in parallel. The absolute value jumped from 0.02 to 0.25, exceeding the instability threshold of 0.1 and triggering an alarm. This data indicates that the parallel operation of total flow control and symmetry diagnosis can identify quality degradation of the spray pattern while maintaining the stability of the total flow rate. Experimental data shows that the dynamic impedance adaptive system of this invention can effectively maintain the stability of the mass flow rate when faced with process disturbances caused by changes in fluid properties and feeding conditions. Furthermore, its symmetry diagnosis mechanism can identify the decline in the quality of the spray spatial distribution caused by abnormal nozzle conditions when the total flow rate is stable, thus providing objective data support for the accuracy and stability of process control.

[0035] To further verify the necessity and non-obviousness of the technical concept of performing total analysis and spatial difference parallel processing of single sensor source information to achieve process diagnosis, the following comparative experiments were conducted.

[0036] Comparative Example 1: This comparative example aims to verify the actual effect of a simplified technical solution that a person skilled in the art might use when attempting to solve the problems raised in the background art. This simplified solution is completely consistent with the test group in Example 2 using the method of the present invention (hereinafter referred to as the test group of the present invention) in terms of test platform, test procedure and target parameters. The essential difference lies in the structure of the electrical sensing device: The electrical sensing device used in this comparative example is an integral non-segmented ring conductive electrode. This structure can obtain the total space charge generated by oil mist friction and generate a single feedback signal that is proportional to the total mass flow rate. However, it does not have the ability to distinguish the spatial distribution differences of the signal in terms of physical structure. The controller of the system also adopts the proportional-integral (PI) control algorithm. Based on the deviation between the single feedback signal and the target value (50.0 g / min), the drive signal of the fuel injector is adjusted in a closed loop. The test process also simulates two working conditions: the baseline working condition and the nozzle partial blockage. The specific test data are recorded in Table 2 below.

[0037] Table 2: Performance comparison test data of the present invention and the simplified scheme under different working conditions.

[0038] Referring to Table 2, under the baseline condition of test number 1, both schemes could achieve a target mass flow rate close to 50.0 g / min through closed-loop feedback. Under the nozzle partial blockage condition of test number 2, the system of comparative example 1 also demonstrated stable control over the total flow rate. Its single ring electrode detected the flow rate decrease trend caused by blockage, and the PI algorithm in the controller automatically increased the drive signal output to the fuel injector, successfully compensating for and maintaining the total mass flow rate at 50.1 g / min. During this process, the system did not output any abnormal alarms. However, the weighing analysis of the receiving substrate after the test showed that there was a difference of more than 40% in the amount of oil coating on the left and right sides of the substrate, and the coating uniformity was severely deteriorated. At the same time, under the same conditions, the execution process control module of the test group of this invention maintained the total mass flow rate at 50.1 g / min, while its execution status diagnosis module monitored the symmetry index in parallel. The absolute value jumped from 0.02 to 0.25, stably exceeding the preset instability threshold of 0.1, and triggered an alarm for asymmetric spray pattern. The test results show that although the simplified technical solution of using a single integral electrode for total flow feedback can solve the problem of total flow fluctuation caused by process disturbance, for faults such as local nozzle blockage that simultaneously cause changes in total flow and spatial distribution distortion, the compensation action of its control loop will mask the physical manifestation of the fault. This blinds the process execution quality problem that seriously affects product quality. The test results prove from the opposite perspective that the technical solution of the present invention, which divides the sensing electrode and performs parallel summation (for total flow control) and differential (for symmetry diagnosis) on the sensing signals from the same source, is the key to ensuring process accuracy and stability.

[0039] Example 3: This example combines Figures 1 to 3 A description of a dynamic impedance adaptive system for non-destructive oil coating of food is provided, such as... Figure 1 As shown, the real-time sensing signals collected by the electrical sensing device are sent to the signal processing module. This module outputs the total flow signal and the spatial distribution signal in parallel. The total flow signal is sent to the execution process control module and compared with the externally input process target value to generate a drive signal to control the fuel injection actuator. At the same time, the spatial distribution signal is sent to the execution status diagnosis module and compared with the diagnostic threshold set by the operator or the human-machine interface (HMI). The execution status diagnosis module also receives the drive signal from the fuel injection actuator and the internal signal from the execution process control module to perform a comprehensive execution status analysis and output the diagnostic results to the operator's HMI.

[0040] like Figure 2 As shown, the horizontal axis represents time in seconds (s), the left vertical axis represents total flow rate in grams per minute (g / min), and the right vertical axis represents the symmetry index. The total flow rate curve, represented by the solid line in the figure, remains stable near the target value of 50.0 g / min under the closed-loop feedback of the execution process control loop. The symmetry index, represented by the long dashed line in the figure, shows a continuous unidirectional increase over time, and finally stably crosses the diagnostic threshold of 0.10, represented by the short dashed line in the figure, at the 92nd second. This indicates that the system's symmetry diagnostic loop successfully identified the deterioration of the spray spatial distribution pattern, which the total control loop itself could not perceive.

[0041] like Figure 3 As shown, the interaction timing between the operator, controller, fuel injector actuator, electrode sensor, and storage module is illustrated. After the operator starts the system, the controller enters engineering debugging mode and sends an initial drive signal to the fuel injector actuator, which then begins fuel atomization, thus initiating a baseline learning process that lasts for 30 minutes. During this period, because the fuel mist carries an electric charge, the electrode sensor continuously sends four sensing signals to the controller. The controller then cyclically calculates the symmetry index and records the drive signal values. After the learning cycle ends, the controller performs statistical analysis on all the collected data, calculates the mean and standard deviation, and generates a baseline value of 5.2V representing the health status of the fuel injector actuator and a symmetry threshold of 0.1 representing the symmetry of the spray pattern. Finally, these two key parameters are stored in the storage module, and a signal indicating that the baseline learning is complete is sent back to the operator.

[0042] Example 4: This example aims to provide a standardized engineering calibration procedure for determining the key control parameters and diagnostic thresholds of the system, providing a basis for adapting various parameters after initial deployment or replacement of core components. In a specific application, after the dynamic impedance adaptive system of this invention is physically installed on a new biscuit production line, the control algorithm parameters and diagnostic model baseline of its internal controller are in the initial state. At this time, in order to adapt the system to the specific working conditions of the current production line, namely the viscosity-temperature characteristics of the specific edible oil used, the pressure characteristics of the gas supply network, and the flow resistance characteristics of the new oil injection actuator, a system initialization and parameter self-tuning operation needs to be performed. The first step of this operation is to tune the parameters of the proportional-integral (PI) control algorithm in the total quantity control loop. The operator enters the engineering debugging mode on the human-machine interface of the controller and adjusts the integral gain of the algorithm. Set the value to zero and temporarily disable the symmetry diagnostic alarm function; then, start the fuel injection actuator and gradually increase the proportional gain. The value is displayed, and the controller's drive signal output waveform is observed using the built-in data oscilloscope function; when Increase to a critical value At this point, the drive signal begins to exhibit sustained constant-amplitude oscillations; the period of this oscillation is then recorded. Seconds; based on the Ziegler-Nichols tuning method, the controller can automatically calculate the tuning parameters applicable to this PI controller, namely the proportional gain. Integral gain This set of parameters is then fixed as the operating parameters under the current working conditions. Through this procedure, the key parameters of the PI control algorithm can be determined based on the dynamic response characteristics of the system.

[0043] After the PI control parameters are tuned, baseline self-learning of the system diagnostic model is performed. The operator installs a brand new clean nozzle on the fuel injection actuator and operates under the baseline condition, i.e., fuel temperature 25°C. At a feeding pressure of 0.4 MPa, the system was started and allowed to run continuously and stably for 30 minutes; during this period, the controller continuously recorded the symmetry index at a sampling rate of 100 Hz. The instantaneous value and the drive signal voltage value output by the PI controller; after a 30-minute self-learning cycle, the controller analyzes all 180,000 symmetry indices collected. Perform statistical analysis on the sample points and calculate their mean. with standard deviation The controller then sets the symmetric instability threshold to... ,Right now The controller takes one significant digit as 0.1. Simultaneously, it performs an arithmetic average of all drive signal voltage samples collected within the same period and stores the resulting average value of 5.2V as a baseline value characterizing the health status of the fuel injection actuator for subsequent long-term statistical trend analysis. By executing the standardized procedures for parameter tuning and baseline self-learning, the key algorithm parameters and diagnostic thresholds within the system are determined through objective measurement and calculation of the current specific operating conditions. At this point, the system has completed its initial configuration, and its control and diagnostic functions operate based on the parameters calibrated for the current production environment.

[0044] Example 5: This example aims to provide specific deployment procedures and coping mechanisms to demonstrate the adaptability and robustness of the technical solution of the present invention in the face of batch differences in production materials and fluctuations in environmental conditions. In a food processing workshop, the production line needs to switch between different batches of biscuit products according to orders. Different batches of biscuits have different surface wettability due to differences in their raw material formulas or baking processes. At the same time, the ambient humidity in the workshop will also fluctuate within a large range due to cleaning operations or weather changes. Both of these factors will affect the measurement accuracy of the active detection device and the compensation effect of the feedforward signal in the specific implementation. To address this, the system is configured to perform an automated pre-calibration procedure before each new production task is started.

[0045] The procedure is first initiated with the conveyor belt unloaded. The ion generator of the active detection device operates continuously for 60 seconds, during which the controller continuously acquires signals from the sensing electrodes to determine the baseline ion current under the current environment. Meanwhile, an industrial-grade humidity sensor integrated into the system control cabinet transmits the real-time measured ambient humidity value, such as 65%RH, to the controller. The controller stores a humidity-baseline current lookup table established during initial system installation through calibration experiments at different humidity levels. Based on the currently measured humidity value, the controller retrieves or interpolates from this lookup table to calculate the corresponding drift-free theoretical baseline ion current value, and uses this theoretical value to correct the measured value. This compensates for the impact of changes in ambient humidity on the measurement. Next, the operator places 10 representative biscuit samples from the batch sequentially onto the conveyor belt and passes them through an active detection device. The controller then calculates the average wettability index of the batch of samples. Finally, the controller, based on the product's process formulation document, specifically addresses this... The gain coefficient corresponding to the value Automatic update feedforward compensation algorithm In Through this two-step automated pre-calibration, the system completes parameter adaptation to environmental changes and material characteristics before entering mass production.

[0046] Example 6: This example aims to provide specific engineering design and implementation procedures for determining the geometric parameters of key hardware in the technical solution of this invention, as well as the fault-tolerant processing mechanism of the system when facing sensor signal anomalies. To determine the specific geometric dimensions of the annular electrode structure in the electrical sensing device to obtain a high signal-to-noise ratio under a specific spray cone shape, finite element analysis (FEM) is used for electric field simulation during the system design phase. The objective function of the simulation model is to maximize the amplitude of the induced signal on the electrode, while maintaining the airflow disturbance of the electrode structure on the atomized oil path within the designed parameters. Under certain limits, by iteratively optimizing three key geometric parameters—the central aperture of the annular structure, the radial width of the electrode, and the axial length—a set of design parameters suitable for spray cone angles in the range of 30 to 45 degrees was obtained. Specifically, the central aperture is 1.5 to 2.0 times the nominal diameter of the spray, and the ratio of the radial width to the axial length of the electrode is between 0.8 and 1.2. Accordingly, for a spray application with a nominal diameter of 50 mm, the specific dimensions of the electrical sensing device were determined as follows: central aperture 90 mm, radial width of the fan-shaped electrode 20 mm, and axial length 20 mm.

[0047] To address edge conditions such as strong electromagnetic interference or sensor hardware failure that may occur in industrial environments, the controller integrates a set of signal validity identification and fault-tolerant processing logic. The controller continuously monitors the raw voltage signal output from each signal amplification circuit at millisecond intervals to determine if it is in one of two preset failure states: signal saturation (determined when the signal voltage exceeds 98% of the system's analog-to-digital converter (ADC) range for more than 100ms, e.g., exceeding 4.9V at a 5V range), or signal loss (determined when the signal voltage falls below a preset value for more than 500ms). The background noise threshold is set to, for example, 0.05V. When the signal of any channel meets any of the above failure conditions, the controller determines that the sensor signal of that channel is invalid and immediately enters the preset fault-tolerant operation mode. In this mode, the total quantity control loop and the symmetry diagnostic loop will ignore the data from the failed channel and perform calculations and control based on the signals of the remaining healthy channels. At the same time, the system will generate a hardware fault alarm with the highest priority to prompt the operator to perform maintenance. If all channel signals fail simultaneously, the system will automatically switch to the safety mode, that is, drive the fuel injection actuator to operate at the average drive signal value of the recent period. Spraying is stopped after 30 seconds to await manual intervention. In a flexible production scenario requiring frequent product variety switching, the production line needs to sequentially apply oil to square cookies (100mm wide) and strip cookies (20mm wide). When producing square cookies, the system does not activate the controllable high-voltage power supply; the spray actuator, under the action of the total quantity control loop, forms a standard wide spray that completely covers the surface of the square cookie. When the production line switches to the strip cookie production task, the operator retrieves the corresponding product formula through the human-machine interface. Based on the formula parameters, the controller instructs the controllable high-voltage power supply connected to the four fan-shaped conductive electrodes. A 1500V DC voltage is output, and the system enters a time-division multiplexing operating mode with a cycle of 10 milliseconds. During the driving cycle of this mode, the 1500V voltage is applied to the electrodes, and the resulting repulsive electrostatic field compresses the coverage width of the oil mist cone to approximately 25mm. In the subsequent sensing cycle, the high-voltage power supply is disconnected, and the controller acquires the sensing signal to maintain a stable mass flow rate. Through this electronic method, the system completes the switching of the spray pattern within milliseconds. Compared with mechanical adjustment, this not only reduces oil waste caused by overspraying but also shortens the downtime adjustment time when changing product types on the production line.

[0048] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0049] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A dynamic impedance adaptive system for non-destructive oil coating of food, comprising an oil spraying actuator, an electrical sensing device, and a controller, wherein the controller controls the oil spraying actuator via a drive signal, characterized in that, The electrical sensing device includes multiple electrically isolated conductive electrodes arranged around the path of the atomized fuel sprayed from the fuel injector. The controller is configured to execute the following rules: Rule a: Multiple real-time sensing signals are acquired through multiple conductive electrodes, wherein each real-time sensing signal is generated by the space charge carried by the atomized oil due to triboelectric charging during the atomization process and induced on the corresponding conductive electrode. Rule b involves continuously comparing the sum of multiple real-time sensor signals with a preset target value representing the target mass flow rate, and adjusting the drive signal to the fuel injection actuator in real time based on the deviation generated by the comparison using a preset control algorithm. Rule c generates a diagnostic signal characterizing the spatial distribution symmetry of atomized oil based on the numerical differences between multiple real-time sensor signals, and uses the diagnostic signal to trigger a preset alarm or control command.

2. The dynamic impedance adaptive system for non-destructive oil coating of food according to claim 1, characterized in that, In rule b, the controller is configured to: integrate or average multiple real-time sensor signals in each control cycle to obtain a sum; and use a proportional-integral control algorithm or a proportional-integral-derivative control algorithm to calculate and update the drive signal for the fuel injector actuator based on the deviation between the sum and the preset target value, wherein the drive signal is a voltage signal or pulse width modulation signal used to control the fluid valve inside the fuel injector actuator.

3. The dynamic impedance adaptive system for non-destructive oil coating of food according to claim 1, characterized in that, The controller is also configured to perform long-term statistical trend analysis on the drive signals output to the fuel injection actuator in rule b; Statistical trend analysis includes calculating the moving average of the drive signal over a pre-defined slow timescale that is longer than the control period; Furthermore, when the moving average value deviates continuously in one direction from a baseline value representing the health status of the fuel injector actuator through learning operations or a pre-set value, a predictive maintenance instruction related to the deterioration of the health status of the fuel injector actuator is generated.

4. The dynamic impedance adaptive system for non-destructive oiling of food according to claim 1, characterized in that, In rule c, the controller is configured to: select at least two conductive electrodes, a first conductive electrode and a second conductive electrode, respectively located on opposite sides of the atomized oil path, from a plurality of conductive electrodes; and perform calculations. To calculate the symmetry index, where, The symmetry index The amplitude of the real-time sensing signal generated by the first conductive electrode. The amplitude of the real-time sensing signal generated by the second conductive electrode is used; and a diagnostic signal is generated when the absolute value of the symmetry index continuously exceeds a preset symmetry instability threshold.

5. The dynamic impedance adaptive system for non-destructive oil coating of food according to claim 1, characterized in that, The system also includes a controllable high-voltage power supply connected to multiple conductive electrodes; and the controller is configured to operate in a time-division multiplexing mode: within one drive cycle, the controllable high-voltage power supply is controlled to apply a preset voltage to multiple conductive electrodes, thereby actively changing the spatial distribution of the atomized oil through the resulting electrostatic field; and in the next sensing cycle, the controllable high-voltage power supply is disconnected, and rules a, b and c are executed.

6. The dynamic impedance adaptive system for non-destructive oil coating of food according to claim 1, characterized in that, The system also includes an active detection device located upstream of the fuel injection actuator. The active detection device includes an ion generator for generating a stable ion wind and a sensing electrode for collecting ions that penetrate the food or are not captured by the food. The controller is also configured to generate a feedforward signal characterizing the wettability of the food surface based on the signal from the sensing electrode. Furthermore, the controller is configured to dynamically adjust a preset target value using the feedforward signal before performing the comparison of rule b.

7. The dynamic impedance adaptive system for non-destructive oiling of food according to claim 1, characterized in that, The electrical sensing device has multiple conductive electrodes, which are four electrically isolated sector-shaped conductive electrodes. The four sector-shaped conductive electrodes together form a ring structure and are coaxially located in the path of the atomized oil.

8. The dynamic impedance adaptive system for non-destructive oiling of food according to claim 1, characterized in that, The system also includes multiple signal amplification circuits, which are connected to multiple conductive electrodes respectively. Each signal amplification circuit is a current-to-voltage amplifier circuit with high input impedance, configured to convert the induced current on the corresponding conductive electrode into a real-time sensing signal in the form of a voltage.

9. The dynamic impedance adaptive system for non-destructive oiling of food according to claim 1, characterized in that, The preset alarm or control command triggered by the diagnostic signal in rule c is specifically: an alarm signal indicating that the fuel injection actuator has a partial blockage, or a safety command to stop the operation of the fuel injection actuator.

10. A dynamic impedance adaptive system for non-destructive oiling of food according to claim 5, characterized in that, The controller is also configured to: perform a fast Fourier transform on the raw waveform data of the real-time sensing signal generated by any conductive electrode during the sensing period to obtain its spectrum; and generate a medium state signal based on the ratio of harmonic energy to fundamental frequency energy in the spectrum, and use the medium state signal to trigger an early warning indicating the deterioration of the chemical properties of the oil.

Images