Fermentation production control system for chilies in red oil

Abstract

The invention relates to the technical field of non-electrical variable regulation, and discloses a red oil chili fermentation production control system, which comprises a logic processing unit, a disturbance frequency modulation module and a torque load execution module, and is characterized in that the logic processing unit outputs a perturbation instruction to the disturbance frequency modulation module; the control torque load execution module superposes perturbation frequency signals under the operation rotating speed, obtains the dynamic phase difference of real-time response torque data relative to driving frequency signals, calculates the effective damping component caused by a controlled viscous medium, corrects the offset error of a physical reference line, further determines the fermentation viscosity state and outputs a control instruction, and finally controls the fermentation viscosity state. Mechanical loss and fluid load are separated through frequency domain phase difference, sensing reference self-calibration is achieved, the problem of adjustment misalignment caused by sensing passivation in the grease corrosion environment is solved, and the consistency of fermentation products is ensured.

Description

Technical Field

[0001] This invention belongs to the field of non-electric variable regulation technology, and in particular relates to a control system for the fermentation production of chili oil. Background Technology

[0002] Currently, control systems for non-electrical variables such as temperature and viscosity are widely used in food production processes. Conventional technical solutions often employ fermentation vessels with heat exchange jackets and mechanical stirring mechanisms. Feedback signals are obtained through temperature sensors deployed inside the vessel, and proportional-integral-differential algorithms are used to adjust the flow rate of the heat exchange medium to achieve tracking and adjustment of the preset fermentation process curve. However, in the fermentation process of chili oil, a multiphase system containing vegetable oil, chili solid particles, and microbial metabolites, the rheological properties of the materials undergo nonlinear evolution with the fermentation process. The oil components and high molecular weight substances such as capsaicin in the materials easily form dynamically growing adhesion layers on the inner wall of the fermentation vessel and the surface of the sensor probe, constructing a thermal resistance shielding band that changes over time. This results in an amplitude deviation between the temperature signal collected from the vessel wall and the actual temperature at the center of the material, and introduces nonlinear phase hysteresis that cannot be eliminated by conventional algorithms.

[0003] When fermentation reaches its metabolic peak, the bubbles generated inside the material introduce a cavitation lubrication effect, producing an apparent viscosity reduction phenomenon that interferes with torque monitoring. This masks the true growth state of the layer adhering to the vessel wall. If heat transfer is forcibly enhanced by increasing the stirring speed, the shear-thinning characteristics unique to non-Newtonian fluids will cause a non-substantial decrease in the drive current, inducing the control system to generate incorrect adjustment criteria. Furthermore, existing technologies mostly focus on improving the redundancy of sensing components, neglecting the physical shielding effect of the oily environment on the sensing components and the mechanical loss and entropy increase process of the actuator as the operating cycle increases. It is difficult to fundamentally solve the feedback distortion problem of the control system in long-term operation. For example, Chinese invention patent CN107252053A discloses a tank fermentation preparation process for red oil broad beans, replacing the traditional pit fermentation tank with a fermentation tank and introducing timed material circulation and preset temperature range management. Essentially, this is a... Static regulation based on prior experience lacks the ability to mine the real-time dynamic response characteristics of the stirring actuator, fails to identify signal drift caused by probe oil film passivation, and struggles to decouple apparent load drops caused by cavitation effects. This causes the regulation logic to fall into a regulation blind zone when the sensing baseline shifts imperceptibly, and lacks the ability to identify physical boundary states. Consequently, the actuator struggles to accurately trigger cleaning pulses or correct heat transfer gains during long-cycle fermentation, resulting in localized material charring or unstable fermentation flavor. Furthermore, since the sensing system inevitably experiences detection passivation during fermentation cycles lasting 15 to 30 days, and the friction coefficient of the drive mechanism shifts over time, the control system struggles to distinguish between the actual changes in the physical state of the controlled object and the degradation of hardware performance. This imperceptible drift of the sensing baseline creates a regulation blind zone for the actuator, ultimately leading to localized material charring or deterioration in the consistency of fermentation product flavor.

[0004] Therefore, how to use the dynamic response characteristics of the actuator to identify the physical boundary state inside the fermenter in real time, and to achieve active compensation for thermal resistance interference and sensing drift without adding physical sensing components, has become the technical problem to be solved by this invention. Summary of the Invention

[0005] This invention provides a chili oil fermentation production control system, comprising:

[0006] Logic processing unit, disturbance frequency modulation module, and torque load execution module;

[0007] The logic processing unit is electrically connected to both the disturbance frequency modulation module and the torque load execution module. The logic processing unit is used to execute the following logic:

[0008] Step 101: Output a perturbation command to the perturbation frequency modulation module to control the torque load execution module to superimpose a perturbation frequency signal with an amplitude of 3.5% of the rated current and a frequency in the range of 12Hz to 15Hz at the current speed, and obtain the real-time response torque data of the torque load execution module.

[0009] Step 102: Extract the dynamic phase difference between the real-time response torque data and the drive frequency signal based on the perturbation frequency signal;

[0010] Step 103, the reference self-calibration step, uses the dynamic phase difference to calculate the effective damping component caused by the controlled viscous medium contained in the torque load execution module, and corrects the physical reference line offset error based on the deviation between the effective damping component and the preset mechanical loss model, thus completing the zero-point self-calibration of the adjustment loop; the preset mechanical loss model is used to store the mapping relationship between the speed of the torque load execution module and the reference torque under no-load conditions.

[0011] Step 104: Under the reference of zero-point self-calibration, determine the fermentation viscosity state based on the corrected effective damping component, and output control commands for adjusting the output torque of the torque load execution module or adjusting the heat exchange power.

[0012] Preferably, when the logic processing unit executes step 102, it performs a fast Fourier transform on the real-time response torque data to extract the phase offset angle at the frequency point of the perturbation frequency signal; and identifies the degree of nonlinear distortion of the phase offset angle relative to the perturbation frequency command to quantify the physical friction attenuation of the transmission chain of the torque load execution module; wherein, the physical friction attenuation is used as a dynamic correction parameter of the preset mechanical loss model to update the physical baseline offset error in step 103.

[0013] Preferably, the logic processing unit is also used to execute the angular momentum echo analysis logic: control the torque load execution module to perform pulse deceleration action and collect the damping attenuation curve of the controlled viscous medium under inertial residual; identify the uniformity of the internal flow field based on the damping attenuation curve, and extend the rotation switching cycle of the torque load execution module when the uniformity is lower than a preset threshold.

[0014] Preferably, the logic processing unit includes a shear sensitivity correction module, used to correct the rheological properties of the effective damping component to obtain the corrected fermentation viscosity. The calculation logic is as follows: ,in, The initial viscosity value is calculated based on the effective damping component, β is the shear thinning sensitivity factor of the controlled viscous medium, and Δω is the rotational speed fluctuation of the torque load execution module during the detection cycle; step 104 is based on the corrected fermentation viscosity. Determine the fermentation viscosity state.

[0015] Preferably, when the logic processing unit identifies a central thickening phenomenon in the material based on the fermentation viscosity state, it outputs an increase command to control the torque load execution module to increase the instantaneous shear rate, thereby disrupting the phase accumulation inside the material.

[0016] Preferably, the system further includes a heat exchange regulation module, which is connected to the logic processing unit; the logic processing unit adjusts the output power of the heat exchange regulation module according to the fermentation viscosity state to compensate for the influence of the thermal resistance adhesion layer formed by the solid particles deposited on the inner wall of the fermentation device on the heat exchange efficiency.

[0017] Preferably, the center frequency of the perturbation frequency signal avoids the mechanical resonance point of the transmission components inside the torque load execution module, so as to ensure that the signal-to-noise ratio of the real-time response torque data is not lower than the preset signal-to-noise ratio threshold.

[0018] Preferably, the logic processing unit performs a baseline self-calibration step once in each fermentation control cycle and stores the calibrated zero-point parameters in a real-time database to construct a performance degradation trend model of the torque load execution module throughout its entire life cycle.

[0019] Preferably, when the logic processing unit detects that the physical baseline offset error exceeds the safety limit deviation, it outputs a warning signal indicating that the signal strength attenuation of the sensing hardware has exceeded the limit.

[0020] Preferably, the system also includes a power supply module, which is electrically connected to the logic processing unit and the disturbance frequency modulation module respectively, to provide a stable power input for the entire system.

[0021] Compared with existing technologies, the chili oil fermentation production control system of this invention has the following advantages:

[0022] 1. In the fermentation production control of chili oil, a thermal resistance coupling factor characterizing the material adhesion state on the inner wall of the fermentation vessel is constructed by cross-dimensional correlation between the driving torque signal of the stirring drive unit and the heat exchange parameter signal of the heat exchange jacket. This upgrades the traditional temperature feedback control to a predictive adjustment based on the physical distribution state of the material. The stirring unit is reused as a virtual sensor to detect the thickness of the thermal resistance shielding layer at the bottom of the vessel, identifying the dynamic thermal resistance shielding layer formed by the accumulation of chili solid particles. Thus, before the temperature at the center of the vessel lags behind, the adhesion layer is actively peeled off by outputting a non-uniform shear command and the heat exchange gain is corrected simultaneously. This eliminates the risk of temperature control overshoot caused by local mass transfer obstruction in the solid-liquid multiphase system, ensuring that the heat transfer during the fermentation process is always in a highly efficient linear range, and avoiding material charring or flavor deterioration caused by local heat accumulation.

[0023] 2. By utilizing the high-frequency current fluctuation characteristic entropy of the stirring drive unit, the system achieves accurate identification and compensation for apparent viscosity fluctuations caused by metabolic gas production. During the peak fermentation period of chili oil, the bubbles produced by microorganisms will generate a cavitation lubrication effect in the material system, resulting in a non-substantial decrease in stirring torque. This pseudo-low viscosity phenomenon often masks the true growth state of the deposited layer on the vessel wall. By extracting the random pulsation characteristics in the high-frequency operating current, the resistance drop component caused by bubble interference is decoupled, and the thermal resistance coupling factor is corrected in real time accordingly. This ensures that the system can accurately perceive the true heat exchange environment of the vessel wall even during the intense biological metabolism stage, eliminating the signal deception of the physical regulation system by the biological reaction process, aligning the triggering timing of the cleaning pulse with the true deposition thickness, and ensuring the purity and reliability of the feedback signal under long-term operation.

[0024] 3. By injecting a specific frequency perturbation excitation into the stirring drive signal and extracting the phase lag angle of the response signal, a self-healing sensing benchmark calibration system is established. Considering that the fermentation environment is corrosive to oils and prone to sensor oil film contamination, the long operating cycle of 15 to 30 days will inevitably cause the probe to experience sensing hysteresis or zero-point drift. By utilizing the difference between mechanical friction damping and fluid viscosity damping in the frequency domain phase fingerprint, the effective load component purely caused by the controlled material is extracted from the total torque signal, thereby identifying systematic errors caused by transmission chain wear or sensor probe passivation. This self-calibration mechanism based on physical response consistency enables the control loop to identify hardware aging and operating condition deviations. Without interrupting production for manual calibration, it ensures that the control logic is always anchored on the real physical benchmark, improving the stability of the system in harsh industrial environments. Attached Figure Description

[0025] Figure 1 This is a flowchart of the self-calibration control logic for perturbation signal analysis in this invention;

[0026] Figure 2 This is a block diagram illustrating the hardware architecture and functional module interaction principle of the system of this invention. Detailed Implementation

[0027] The technical solutions of the embodiments of this application will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.

[0028] It should be noted that all directional and positional terms used in this invention, such as: up, down, left, right, front, back, vertical, horizontal, inner, outer, top, bottom, transverse, longitudinal, center, etc., are only used to explain the relative positional relationship and connection between components in a specific state (as shown in the accompanying drawings). They are only for the convenience of describing this invention and do not require that this invention be constructed and operated in a specific orientation. Therefore, they should not be construed as limiting this invention. In addition, the descriptions of "first," "second," etc., in this invention are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated.

[0029] In the description of this invention, unless otherwise explicitly specified and limited, the terms installation, connection, and linking should be interpreted broadly. For example, they can refer to fixed connections, detachable connections, or integral connections; they can refer to mechanical connections; they can refer to direct connections or indirect connections through an intermediate medium; they can refer to the internal connection of two components. For those skilled in the art, the specific meaning of the above terms in this invention can be understood according to the specific circumstances.

[0030] In the description of this specification, references to the terms "an embodiment," "some embodiments," "illustrative embodiments," "examples," "specific examples," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example, and the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0031] This invention discloses a chili oil fermentation production control system. The logic processing unit processes real-time response torque data from the torque load execution module and heat exchange parameter signals from the heat exchange jacket to establish a correlation between mechanical and thermodynamic parameters, thereby regulating non-electrical variables inside the fermentation vessel. During chili oil fermentation, chili flakes easily accumulate on the inner wall of the fermentation vessel, forming a dynamically growing thermal resistance layer, causing a lag in the temperature control loop. To address this challenge, the logic processing unit collects the output torque τ of the drive motor, motor power P, stirring speed ω, and the temperature difference ΔT between the inlet and outlet of the heat exchange jacket from the torque load execution module, and calculates the thermal resistance coupling, which characterizes the degree of heat exchange obstruction at the vessel wall. The thermal resistance coupling factor η is calculated by setting a lower limit of 1.5 rad / s for the speed term. When the real-time measured stirring speed is lower than 1.5 rad / s, the calculation logic automatically forces 1.5 rad / s as the denominator to eliminate numerical divergence caused by the stirring mechanism switching at low speeds or starting and stopping. Simultaneously, the physical essence of this thermal resistance coupling factor is the heat transfer efficiency per unit mechanical power. The logic processing unit collects power P and speed data every 100 ms and stores them in a 50-depth annular buffer. The arithmetic mean of this buffer is used in the calculation to eliminate instantaneous torque electromagnetic noise caused by heterogeneous material collisions. The calculation formula is as follows: Where η is the thermal resistance coupling factor; ΔT is the temperature difference between the inlet and outlet of the heat exchange jacket, in °C; P is the power of the stirring motor, in W; and ω is the stirring speed, in rad / s. The temperature difference ΔT is obtained by temperature sensors deployed at the inlet and outlet of the heat exchange jacket. The power P and the speed ω are read by the internal register of the frequency converter. When the thermal resistance coupling factor η exceeds the preset threshold, it is determined that the increased thickness of the adhesion layer on the reactor wall leads to an increase in heat exchange damping. At this time, the logic processing unit outputs a non-uniform shearing command to drive the stirring unit to alternate between the first preset frequency and the second preset frequency. The shear stress generated by the fluid inertia peels off the adhesion layer. The system synchronously corrects the adjustment gain of the heat exchange medium flow rate in the heat exchange jacket, so that the heat transfer during the fermentation process is in the linear range.

[0032] Establishing the perturbation frequency signal When determining the amplitude parameter, the logic processing unit executes the excitation gain linearity verification procedure, controls the disturbance frequency modulation module to add micro-disturbance components in incremental steps at 0.5% on the current operating current reference, and monitors the characteristic spectrum energy distribution in the feedback current of the torque load execution module in real time. Considering that when the superimposed amplitude is less than 2.5% of the rated current, electromagnetic noise interference caused by the inverter dead zone compensation effect will cause the extraction deviation of the phase lag angle θ to exceed 15.0%, when the amplitude is increased and stabilized at 3.5%, the signal-to-noise ratio of the characteristic component reaches more than 22dB and the root mean square value of the phase calculation residual converges to within 0.005rad. This process establishes the sampling accuracy boundary of the sensing loop in the grease corrosion environment. Specifically, the logic processing unit has a built-in 13.5Hz center frequency, A 2.0Hz digital bandpass filter is used to extract the perturbation response component from the original current signal containing high-order harmonics. To ensure that the calculated residual of the phase lag angle θ converges to within 0.005 rad, the system requires continuous sampling of 8 complete perturbation cycles (approximately 600 ms). The time difference between the current waveform and the original command voltage waveform is identified through a zero-crossing detection algorithm. This time difference divided by the total duration of the perturbation cycle is mapped to the phase angle. The fermentation environment is corrosive to oils, which easily causes oil film contamination on the sensor probe surface, resulting in phase deviation of the sensing signal. The logic processing unit executes a reference self-calibration procedure to correct the physical baseline. The logic processing unit outputs a perturbation command to the perturbation frequency modulation module, controlling the torque load execution module to superimpose the perturbation frequency signal at the current operating speed. perturbation frequency signal The amplitude is set to 3.5% of the rated current, and the frequency range is selected between 12Hz and 15Hz to avoid the mechanical resonance point of the stirring shaft drive assembly. The system uses a digital bandpass filter to extract the perturbation frequency signal from the feedback current. Torque ripple component at the same frequency The logic processing unit calculates the torque ripple component. Relative to perturbation frequency signals The phase lag angle θ is calculated, and the effective damping component induced by the controlled viscous medium is calculated. The calculation formula is as follows: ,in, The effective damping component is expressed in N·m. θ represents the torque ripple component, in N·m; θ represents the real-time phase lag angle, in rad. The no-load reference phase, in rad, is stored for the preset mechanical loss model. The phase lag angle θ is compared with the reference phase. The system identifies the degradation in sensing performance caused by probe oil film contamination based on the offset. The logic processing unit uses a phase lead compensator to correct the temperature signal and complete the zero-point self-calibration of the adjustment loop.

[0033] The system executes an unloaded calibration sequence to establish a mechanical loss model, controls the torque load execution module to run in 5r / min steps within the range of 5r / min to 80r / min, and records the reference torque under equilibrium state at each speed node. Through formula Calculate the reference damping coefficient Where ω is the angular velocity corresponding to the stirring speed; the logic processing unit uses a linear interpolation algorithm to fit each discrete calibration point into a continuous loss compensation curve. During the self-test phase before the start of the fermentation cycle, it monitors the output residual of the motor in a static state. If the residual value exceeds 0.2% of the rated torque, the deviation value is used as the zero-point calibration of the physical baseline offset error. It distinguishes between the mechanical loss entropy increase caused by long-term operation of the transmission components and the load caused by material viscosity damping; for the dynamic balance offset of the stirring shaft system caused by material adhering to the wall during fermentation, the system executes the spectrum obstacle avoidance calibration procedure to adjust the perturbation frequency signal. The frequency band selection criteria were verified. The logic processing unit drove the torque load execution module to perform a frequency sweep operation from 5Hz to 25Hz, collected the mechanical vibration acceleration at the bearing support, and identified the first-order bending resonance peak of the system at 8.5Hz and the transmission chain meshing noise at 18.2Hz. Based on the trade-off of avoiding mechanical resonance points to extend hardware life, the system selected 12Hz to 15Hz as the perturbation frequency signal. Its only operating frequency band.

[0034] Establish perturbation frequency signal At the center frequency, the logic processing unit drives the torque load execution module to perform a frequency sweep operation from 5Hz to 25Hz under no-load conditions. Vibration spectrum is collected by a mechanical vibration acceleration sensor deployed at the bearing support, identifying the first-order bending resonance peak at 8.5Hz and the transmission chain meshing noise frequency band at 18.2Hz. The operating frequency band for this signal is selected as 12Hz to 15Hz. After superimposing perturbation instructions, the signal-to-noise ratio of the characteristic components is not less than 22dB, ensuring that the residual of the dynamic phase difference calculation of the real-time response torque data relative to the drive frequency signal converges within 0.005rad, supporting the sampling accuracy of the sensing benchmark self-calibration procedure under grease corrosion. During the peak fermentation period, microbial metabolism produces bubbles, creating a cavitation lubrication effect in the material system and triggering apparent viscosity reduction. The logic processing unit employs a decoupling mechanism based on high-frequency characteristic entropy to collect the instantaneous current signal of the stirring drive unit at a frequency of 1kHz and extract the fluctuation component from the high-frequency operating current signal. Calculate the fluctuation characteristic entropy using a sliding time window The system generates a viscosity correction coefficient δ based on the mapping relationship. The calculation formula is as follows: Where δ is the viscosity correction coefficient, which is a dimensionless parameter; β is the preset phase interference constant; To calculate the fluctuation characteristic entropy, the logic processing unit incorporates the correction coefficient δ into the calculation logic of the thermal resistance coupling factor η, compensating for the torque drop caused by bubble lubrication and restoring the thermal resistance state of the substrate layer adhering to the vessel wall. The logic processing unit also acquires the instantaneous current signal from the stirring drive unit at a frequency of 1 kHz, selects a 500 ms sliding sampling window, and calculates the probability distribution density of the signal amplitude within the window. Perform numerical calculations based on the information entropy algorithm to calculate the fluctuation characteristic entropy. ,in The probability that the signal amplitude falls within the i-th quantization interval; using the formula The viscosity correction coefficient δ is determined, where β is a preset phase interference constant, which characterizes the torque drop slope of the material system under a specific gas production rate. The viscosity correction coefficient δ participates in the weighted calculation of the thermal resistance coupling factor η, which removes the random resistance pulsation caused by microbial metabolic gas production from the total driving torque, so that the calculation result points to the thermal resistance generated by the adhesion layer on the inner wall of the fermenter.

[0035] To identify the phase inhomogeneity caused by the heterogeneous characteristics of the material, the logic processing unit executes angular momentum echo analysis logic, the system control torque load execution module performs pulsed deceleration, and collects the angular momentum echo components of the stirring drive unit after the deceleration. The system calculates the correlation coefficient of material uniformity. The calculation formula is as follows: ,in, The uniformity correlation coefficient; The echo component of the angular momentum is expressed in N·m·s. Δt is the start time of the pulse deceleration action, in seconds; Δt is the preset monitoring duration, set to 2 seconds; the specific acquisition and conversion logic of the momentum angular echo component Me with composite dimensions is as follows: the logic processing unit uses a dynamic torque sensor to continuously extract the instantaneous mechanical torque feedback signal τ(t) of each frame during the pulse deceleration action (from the start time t0 to the end of the monitoring duration Δt) at a constant sampling rate of not less than 10Hz. Then, through the numerical integration module inside the microprocessor, the time-domain definite integral operation is performed on the torque sequence data in the discrete time domain, thereby converting the directly measured single-dimensional mechanical parameter into a comprehensive momentum angular parameter of N·m·s required for calculation in real time, which constitutes the underlying data input support for subsequent uniformity evaluation, where η is the thermal resistance coupling factor; if the uniformity correlation coefficient If the distribution characteristics exceed a preset threshold, it is determined that there is phase accumulation in the internal flow field. The logic processing unit adjusts the rotation switching frequency of the stirring unit to disrupt the phase isolation inside the material by generating a non-equilibrium signal. Chili oil exhibits shear thinning characteristics in the later stages of fermentation. To prevent non-uniform shearing action from being interfered with by non-Newtonian fluid properties, the system adopts a characteristic step detection procedure. The logic processing unit controls the stirring drive unit to step according to a preset rotation speed sequence. Run and obtain the driving torque corresponding to different timing frequencies. The logic processing unit calculates the shear sensitivity factor λ, using the following formula: Where λ is the shear sensitivity factor; τ is the driving torque, in N·m; The rotational speed is expressed in r / min. The logic processing unit uses the shear sensitivity factor λ to correct the calculation weight of the thermal resistance coupling factor η. When the shear thinning component is detected in the torque decrease, the system extends the duration of the pulse shearing action to improve the system's adaptive range to the evolution of material rheological properties.

[0036] Example 1: When the system faces continuous production of chili oil with a high solid-liquid ratio and a fermentation cycle exceeding 20 days, the accumulated chili flakes and oil mixture on the inner wall of the fermentation vessel form a dynamic adhesion layer with a thickness exceeding 15mm. This leads to a decrease in the heat transfer efficiency between the material in the central area and the heat exchange jacket. The logic processing unit collects the temperature difference ΔT from the temperature sensors deployed at the inlet and outlet of the heat exchange jacket, and simultaneously reads the real-time power P and speed ω of the drive motor from the frequency converter register to calculate the thermal resistance coupling factor η. The formula for calculating the thermal resistance coupling factor η is as follows: Where η is the thermal resistance coupling factor; ΔT is the temperature difference between the inlet and outlet of the heat exchange jacket, in °C; P is the power of the stirring motor, in W; and ω is the stirring speed, in rad / s. The product of power and speed in the denominator essentially represents the intensity of the comprehensive work output by the drive motor to overcome the boundary friction of the high-viscosity adhesion layer. Before executing the above calculation, the logic processing unit has a built-in dimensionless processing mechanism for the input variables. That is, it extracts the pure numerical scalar parameters of P and ω measured in real time under their respective international standard units and participates in the multiplication operation. Then, the calculation result η is transformed into a pure empirical coefficient that only reflects the relative trend of local heat transfer deterioration. This avoids the absolute mathematical mismatch of macroscopic mechanical dimensions in the extraction process of this specific characteristic parameter. When the thermal resistance coupling factor η continues to rise to 0.92, the logic processing unit is triggered to output a micro-perturbation frequency signal with a frequency of 14.5Hz and an amplitude of 3.5% of the rated current to the perturbation frequency modulation module. The system extracts the torque fluctuation component from the real-time response torque data. The effective damping component is calculated by combining the phase lag angle θ. The calculation formula is as follows: ,in, The effective damping component is expressed in N·m. θ represents the torque ripple component, in N·m; θ represents the real-time phase lag angle, in rad. To store the no-load reference phase in rad for the preset mechanical loss model, and to achieve precise injection of the aforementioned perturbation frequency signal into the high-inertia torque load execution module, the system's underlying perturbation frequency modulation module avoids the voltage-frequency ratio control strategy of conventional frequency converters, instead adopting the torque-current decoupling mechanism in the vector control architecture. The perturbation command generated by the logic processing unit is configured as a feedforward additional AC component, directly bypassing the speed loop and injected into the quadrature-axis torque current setpoint of the motor core controller. This allows the frequency converter to maintain a constant direct-axis excitation current while forcibly generating a pure electromagnetic torque ripple equivalent to 3.5% of the rated current using its internal high-frequency current closed-loop regulator. This completely overcomes the low-pass attenuation effect of stator inductance on high-frequency signals, ensuring the hardware execution capability of perturbation excitation. To support the aforementioned physical superposition process, this system independently encapsulates a set of high-resolution digital-to-analog converter arrays and digital waveform generators within the perturbation frequency modulation module. The generator synchronously parses the digital wavetable file issued by the logic processing unit through the controller LAN bus and translates it into an analog AC micro-level signal. Subsequently, it relies on a dedicated isolation operational amplifier to enhance the load driving capability and finally feeds it directly to the high-speed analog feedforward input hard interface reserved on the inverter drive motherboard through hard-wired routing, thereby establishing an accurate electrical signal injection link to resist electromagnetic pulse interference.

[0037] The system identifies that the current torque increase signal contains 78% contribution from wall adhesion and 22% contribution from sensor probe oil film shielding. The logic processing unit controls the torque load execution module to switch between a first preset frequency of 35Hz and a second preset frequency of 50Hz, executing a non-uniform pulsed shearing action. This utilizes the shear stress generated by fluid inertia to peel off the adhesion layer on the reactor wall, causing the thermal resistance coupling factor η to drop back to 0.45 within 120s. During the peak fermentation stage involving microbial metabolic gas production, bubbles generated inside the material create a cavitation lubrication effect in the material system, leading to apparent viscosity reduction. If the regulating circuit directly receives the torque decrease signal at this time, it will cause a false action to reduce the heat transfer intensity, inducing local overheating inside the reactor. This technical solution extracts the high-frequency fluctuation component by collecting the instantaneous current signal of the drive unit in the torque load execution module. The fluctuation characteristic entropy is calculated using a sliding time window of 500ms. When the characteristic entropy of fluctuation When the system reaches 1.8 times the preset threshold, the logic processing unit performs nonlinear gain compensation on the thermal resistance coupling factor η using the viscosity correction coefficient δ. The formula for calculating the viscosity correction coefficient δ is as follows: Where δ is the viscosity correction coefficient; β is the preset phase interference constant; As the characteristic entropy of fluctuation, the system decouples the random pulsations caused by the collapse of metabolic bubbles from the physical damping model, identifies the true viscosity state of the material, and maintains the temperature control error within ±0.2℃ by keeping the heat exchange medium flow rate at 2.5 m³ / h.

[0038] To address the non-Newtonian fluid shear thinning characteristics that occur in the later stages of material processing, the logic processing unit obtains the rotational speed step sequence by executing a feature step detection procedure. corresponding driving torque The calculated shear sensitivity factor λ is 0.65. The formula for calculating the shear sensitivity factor λ is as follows: Where λ is the shear sensitivity factor; τ is the driving torque in N·m; and n is the rotational speed in r / min. When performing this correction evaluation, the logic processing unit, based on the inherent isomorphism of the viscoelastic rheological response, directly assigns the macroscopic dynamic characterization factor λ, calculated based on the step probe sequence, to the data register node containing the shear thinning sensitivity factor β used to correct the effective damping component via its internal data redirection bus. This allows the uniformly calculated sensitivity value to synchronously participate in the multi-dimensional logic branch calculations for cavitation apparent viscosity reduction decoupling and viscosity characteristic state correction. The logic processing unit adjusts the duration constant of the pulse shear action according to this factor, enabling the system to adjust the duty cycle of the high-frequency shear pulse based on the instantaneous rheological changes of the material. When performing the cleaning action, the system dynamically correlates the cleaning frequency with the shear stress response characteristics of the material, correcting the uneven heat transfer caused by material stratification. To verify the consistency of the final product preparation, after fermentation, the system controls the torque load execution module to perform a pulsed deceleration action and collects the angular momentum echo component of the stirring drive unit after the deceleration action. Calculate the correlation coefficient of material uniformity. Uniformity correlation coefficient The calculation formula is as follows: ,in, The uniformity correlation coefficient; The echo component of the angular momentum is expressed in N·m·s. Δt is the start time of the pulse deceleration action, in seconds; Δt is the preset monitoring duration, set to 2 seconds; η is the thermal resistance coupling factor. The start time t0 in this calculation system does not refer to the absolute clock coordinate from the system's power-on operation or the start of the entire fermentation cycle, but rather to the relative time parameter recalculated by the logic processing unit after clearing the local high-frequency timer through an internal soft interrupt mechanism at the instant each pulse deceleration detection action is triggered. Since each pulse excitation detection action forces the local initialization of the time reference axis, this time decay weighting factor effectively eliminates the invalid time accumulation distortion effect brought about by the entire life cycle of the fermentation process. In this embodiment, by adjusting the rotation switching frequency of the stirring unit, the material uniformity correlation coefficient is... Ultimately, the content of capsaicin and the peroxide value of oil in each layer of the material in the fermentation vessel were sampled and tested. The standard deviation of their distribution was less than 3% of the average value, indicating that the material system had reached the preset phase distribution accuracy. In addition, the judgment boundaries and mandatory lower limits of the above-mentioned non-Newtonian rheological behavior were all derived from the orthogonal experimental parameter calibration performed offline in advance. Specifically, the results were obtained by using a rotational rheometer to perform a logarithmic sweep frequency test on the background material of red chili oil from 0.1 rad / s to 10 rad / s. The experimental spectrum clearly revealed that when the rotation speed was below 1.5 rad / s, the strong initial yield stress inside the material would cause the calculated denominator to approach zero drastically, thus inducing data overflow and divergence. The dimensionless reference mean domain of the shear sensitivity factor λ was also directly derived from the physical property constant by rigorously calculating the slope of the double logarithmic coordinate of the rheological curve in the shear-thinning interval.

[0039] Example 2: This verification experiment was conducted on a 500L pilot-scale fermentation vessel platform. The data originated from real-time operating parameters collected by the physical experiment platform. The platform was equipped with a temperature sensor with a measurement accuracy better than 0.1℃ and a dynamic torque sensor with a resolution of 0.05 N·m. The logic sampling frequency of the control system was set to 10Hz, and the sampling rate for high-frequency feature extraction of motor current was set to 1kHz. The decision logic for setting the excitation signal strength was to balance the sensing resolution with the mechanical system load, i.e., when the perturbation frequency signal... When the amplitude is set to 3.5% of the rated current, a characteristic response with a signal-to-noise ratio greater than 15dB is generated, while avoiding mechanical fatigue damage to the motor drive chain. In order to simulate industrial electromagnetic environment interference, Gaussian white noise with a signal-to-noise ratio of 20dB is actively superimposed during the test.

[0040] The control system consists of the sample group of this invention, the control group, and a sample group with partially missing features. The control group uses conventional PID isothermal control, while the sample group with partially missing features is stripped of effective damping components. The reference self-calibration logic was used. During the experiment, a semi-cured grease film with a thickness of 0.2 mm was applied to the surface of the sensor probe to simulate the sensing passivation condition. Measurement data showed that in the sample group lacking self-calibration logic, the phase lag of the sensing probe drifted from 12.5 ms to 48.2 ms, causing the system to misjudge the increase in thermal resistance caused by the grease film as material adhesion, thereby triggering unnecessary pulse shearing action. In the sample group of this invention, the system identified the sensing reference offset by the offset of the phase lag angle θ and calculated the effective damping component. The value stabilized at 4.25 N·m, with a deviation rate of less than 1.0% from the no-load calibration value of 4.21 N·m. This effectively shielded the control loop from interference caused by hardware performance degradation. Although the change in the pure physical friction background caused by the wear of the transmission bearing is essentially a broadband energy dissipation phenomenon covering the entire frequency band, considering that the mechanical chain exhibits a strict amplitude-frequency linear translation response mechanism for the damping transmission attenuation characteristics of the external excitation signal in the high signal-to-noise ratio flat frequency band far from the first-order resonant natural frequency and the higher-order meshing point, the system only needs to extract the phase offset mapping value of a single point within this specific narrow safety frequency window. It can then use the inherent transmission constant of the rigid system to calculate the global baseline offset of the entire transmission chain surface without error. This confirms the synergistic effect between phase feature extraction and thermal resistance calculation logic.

[0041] In this synergistic process, although the micron-level oil film passivation on the surface of the local thermal sensing probe does not possess the physical ability to directly change the macroscopic mechanical damping of the fermenter's main shaft, the temperature probe and the stirring transmission assembly are exposed to the same oil corrosion and solid phase accumulation environment for a long time. The system's underlying model determines that the phase accumulation rate on the probe surface and the viscous wear increment in the transmission chain gap maintain a highly nonlinear synchronization on the physical evolution time axis. Therefore, the logic processing unit essentially uses the macroscopic transmission chain physical friction attenuation extracted from the perturbation frequency signal as an equivalent environmental probe. By consulting the attenuation trend mapping table pre-stored in the real-time database, it indirectly deduces the signal hysteresis drift that occurs in the local thermal probe, thereby establishing a causal-free logical bridge between microscopic thermal resistance and macroscopic torque. Active compensation of the hardware baseline was achieved without violating the physical principle of cross-scale operation. The logic processing unit directly converted the time drift of the hardware baseline, which deteriorates over time, extracted from the mapping table into the lead time constant inside the phase lead compensator. Then, a first-order differential feedforward compensation component, which strictly corresponds to this time constant, was dynamically superimposed in the first-order inertial filter adjustment algorithm. The calculus response characteristics were used to predict and dynamically align the hysteresis temperature feedback level under the oil film, thereby inversely offsetting the interface heat conduction time delay caused by physical shielding at the control signal flow level. To address the viscosity reduction artifact caused by microbial gas production, three gradient control environments with low, medium, and high gas production rates were set up in the experiment. The measurement data showed that as the gas production rate increased regularly, the fluctuation characteristic entropy increased. The viscosity coefficient η exhibits a non-linear increasing trend from 1.12 to 3.85. Without viscosity correction factor δ compensation, the torque feedback value of the control group drops to 18.5% of the rated torque, inducing the temperature control loop to erroneously reduce the heat transfer intensity. However, in the sample group of this invention, after performing gain compensation on the thermal resistance coupling factor η through the correction factor δ, the zero-point drift of the physical baseline under bubble disturbance is limited to within 2.8%. When the bubble volume fraction exceeds the saturation threshold of 30%, the calculated viscosity correction factor δ enters the saturation compensation region, confirming the stability of this method under extreme fermentation peak conditions, establishing the optimal working window for determining the thermal resistance coupling factor η, and determining the correlation coefficient of material uniformity at the end of the experiment. Echo analysis of the angular momentum showed that the measured values ​​in five consecutive production batches ranged from 0.042 to 0.048, which is lower than the critical threshold of 0.15 for determining flow field imbalance. By calculating the shear sensitivity factor λ and performing characteristic step detection, the system identified shear thinning characteristics in the later stages of fermentation, and the calculated values ​​were... The mean value was 0.68. Based on this, the pulse shearing duration after dynamic extension ensured the suspension stability of chili flakes in the reactor. Sampling and testing results showed that the concentration difference of capsaicin at different depths in the fermentation reactor decreased from 15.6% in the control group to 2.4%, and the standard deviation of the distribution of oil peroxide value decreased from 0.12 to 0.03. This proves that the technical solution of deep coupling between physical feature sensing and the action intensity of the actuator can achieve the goal of uniform control of multiphase system.

[0042] Example 3: This example combines Figures 1 to 2 A description of a fermentation production control system for chili oil, such as... Figure 1 As shown, in step 101, a perturbation command is output to the perturbation frequency modulation module to control the torque load execution module to superimpose a perturbation frequency signal with an amplitude of 3.5% of the rated current and a frequency in the range of 12Hz to 15Hz at the current speed, and to obtain the real-time response torque data of the torque load execution module. In step 102, the dynamic phase difference between the real-time response torque data and the drive frequency signal is extracted based on the perturbation frequency signal. Then, in step 103, the effective damping component induced by the controlled viscous medium is calculated using the dynamic phase difference, and the physical baseline offset error is corrected based on the deviation between the effective damping component and the preset mechanical loss model to complete the zero-point self-calibration of the adjustment loop. Finally, in step 104, under the reference of the zero-point self-calibration, the fermentation viscosity state is determined according to the corrected effective damping component, and a control command for adjusting the output torque of the torque load execution module or adjusting the heat exchange power is output.

[0043] like Figure 2As shown, the system is centered on a central logic processing unit, i.e., an industrial control host or embedded controller. It is connected to a power supply module via a power input to power the system, and is also connected to a perturbation frequency modulation module to generate micro-perturbation frequency signals and commands, a shear sensitivity correction module to perform rheological property correction and viscosity calculation, a angular momentum echo analysis logic to identify internal flow field uniformity, a real-time database for storing reference self-calibration parameters and performance degradation trends, a mechanical loss model, and a heat exchange regulation module to compensate for the influence of thermal resistance adhesion layers. This central logic processing unit receives sensor signals from field sensing terminals. The terminal includes inlet and outlet temperature sensors for the heat exchange jacket, a mechanical vibration acceleration sensor, and a real-time phase and torque monitoring probe. Simultaneously, the central logic processing unit sends perturbation and control commands to the torque load execution module, i.e., the variable frequency motor drive assembly, driving it to receive perturbation frequency and speed commands and execute pulse-type deceleration or rotation direction switching actions. This, in turn, feeds back real-time response torque data to the central logic processing unit to form a torque and phase data stream. In addition, the central logic processing unit sends heat exchange power adjustment commands to the heat exchange device, i.e., the fermenter heat exchange jacket, controlling it to perform actions to adjust the heat exchange power to cope with the fermentation state under the controlled viscous medium thermal resistance environment.

[0044] Example 4: During the self-check phase before the system starts fermentation production, the fermentation vessel is emptied. To establish the background physical damping characteristics of the drive mechanism, the logic processing unit initializes the parameters of the mechanical loss model by executing the benchmark self-calibration procedure. The logic processing unit outputs a perturbation command to the perturbation frequency modulation module. The drive torque load execution module performs a step-by-step frequency sweep action within the speed range of 30 r / min to 60 r / min. The logic processing unit collects the phase lag angle θ corresponding to each speed node in real time and stores it in the mechanical loss model. At a speed of 45 r / min, a perturbation frequency signal of 13.5 Hz is superimposed. Under operating conditions, the measured real-time phase lag angle under no-load conditions is 0.12 rad. The system sets this value as the globally unique no-load reference phase. This procedure directly extracts the background damping generated by the friction of the transmission chain through the physical no-load response, and corrects the sensing baseline; the system calibrates the thermal resistance sensing baseline to determine the thermal resistance coupling factor. The trigger threshold was determined by injecting 200L of pure vegetable oil into the fermenter as a reference medium, turning on the heat exchange jacket, and adjusting the stirring unit to run at 40r / min. The temperature difference ΔT between the inlet and outlet of the heat exchange jacket stabilized at 5.2℃. The motor power P read from the inverter register was 1250W. The logic processing unit calculated the reference thermal resistance coupling factor in the pure oil phase. The threshold value is 0.0416. Based on the physical parameters of the reference medium, the system sets the trigger threshold of the thermal resistance coupling factor η to be 0.0416. When the deposition of chili flakes during fermentation causes the calculated η to exceed 0.0624, the system detects the formation of a heat transfer barrier and initiates a non-uniform shear cleaning action. Through calibration based on known physical property references, the system transforms the control decision into online measured physical ratios. In this judgment logic, when the thickness of the adhesion layer on the fermenter wall increases, causing an internal heat transfer barrier, the effective heat exchange between the heat exchange jacket and the fermenter should be reduced. However, due to the extremely high local velocity gradient formed by the high-viscosity non-Newtonian fluid adhesion layer in the near-wall micro-region, the mechanical force generated when the stirring blades scrape the wall surface... Shear friction dissipation increases sharply; the mechanical energy originally used to drive fluid circulation is directly and transiently converted into a large amount of frictional heat at the interface between the thermal resistance layer and the vessel wall, and is preferentially and forcibly absorbed by the cooling medium of the heat exchange jacket that is close to the vessel wall. This results in an abnormally high outlet temperature of the heat exchange medium flowing through this area, causing the apparent inlet and outlet temperature difference ΔT of the jacket itself to exhibit a counterintuitive surge under the condition of local heat exchange obstruction. The calculated coupling factor η rises sharply accordingly. The system relies on this correlation effect dominated by boundary shear heat generation to identify the thickness change of the adhesion layer.

[0045] To compensate for cavitation lubrication during peak microbial gas production, the logic processing unit calibrates the phase interference constant β. Under conditions where the material in the fermenter is at normal viscosity and there is no biological metabolic gas production, a controlled injection of air pulses at a flow rate of 0.5 L / min is introduced into the bottom layer of the material to simulate the microbial gas production process. The logic processing unit acquires the instantaneous current signal of the stirring drive motor and extracts the high-frequency fluctuation component. And calculate the wave characteristic entropy Measurement data show that when the bubble volume fraction is in the low load range of 5%, the fluctuation characteristic entropy... The effective damping component induced by the controlled viscous medium is 1.25. A 3.2% drop occurred. The logic processing unit calculated the disturbance constant β based on the principle of conservation of physical quantities. The calculation formula is as follows: Where β is the phase disturbance constant and δ is the viscosity correction coefficient; The value is the fluctuation characteristic entropy. In this embodiment, β is calculated to be 0.026. The system stores this value in the register of the logic processing unit for physical reconstruction of the sensed signal during the subsequent fermentation peak. The above-mentioned correlation formula characterizing the nonlinear coupling reciprocal relationship between the fluctuation characteristic entropy and the viscosity drop effect is derived from the extraction of the multi-configuration confined fluid dynamics test model in the offline calibration program of this equipment. The logic processing unit pre-reads a large amount of bench torque test array data for equal-proportion red chili oil matrix under different gradient standard air inflow and determined speed combinations. Using the internally integrated least squares curve fitting engine, it performs regression iteration calculations on the microscopic high-frequency electromagnetic current envelope energy and the macroscopic apparent load relative drop amplitude until it approaches the physical limit convergence boundary and then stores the compensation algorithm path. When fermentation enters the metabolic peak period on the 15th day, the system monitors the fluctuation characteristic entropy. Upon reaching step 2.45, an increase in internal bubble concentration was detected. The logic processing unit generated a viscosity correction coefficient δ of 0.94 based on a preset β value and performed a weighted correction on the real-time calculated thermal resistance coupling factor η. The correction results showed that the thermal resistance coupling factor η, reflecting the adhesion state to the vessel wall, remained stable at 0.052, not reaching the cleaning threshold. By maintaining a stable heat exchange power, the system avoided reducing the heat exchange intensity due to bubble signal interference. After fermentation, the system performed angular momentum echo analysis to calculate the generated uniformity correlation coefficient. The value is 0.045. After sampling and testing, the consistency deviation of the phase distribution of the material is kept within 2.5%.

[0046] Example 5: In the offline calibration scenario before the new equipment is put into production, the fermenter is in an emptied state, and the system drive motor runs in 5r / min increments within the range of 5r / min to 80r / min. Each speed node runs for 300s to allow the bearing lubrication film to reach a balanced state. The logic processing unit records the reference current component and reference torque feedback value at each speed node. Calculate the reference damping coefficient corresponding to each rotational speed ω. The calculation formula is as follows: ,in, The reference damping coefficient; The reference torque feedback value is in N·m; ω is the rotational speed in rad / s; the logic processing unit uses an interpolation algorithm to fit the discrete calibration points into a continuous loss compensation curve, and reads the output residual under the static state of the motor before each production cycle starts. If the measured zero point deviates from the rated torque by more than 0.2%, zero recalibration is performed. This process removes the signal fluctuations caused by the wear of the transmission components from the perception model.

[0047] When the system is deployed in fermentation units with different heat exchange areas, 300L of calibration water is injected into the fermenter and the inlet water pressure of the heat exchange jacket is maintained at 0.3MPa. The flow rate of the heat exchange medium is adjusted to perform a step change within the range of 1.0m³ / h to 4.0m³ / h. The logic processing unit monitors the dynamic response curves of the temperature rise rate inside the fermenter and the inlet and outlet temperature difference ΔT, and calculates the thermal time constant of the current execution unit based on a first-order inertial plus hysteresis model. and heat exchange gain coefficient The system logic is then adjusted to control the parameters of the current fermentation unit after the thermal resistance coupling factor η is adjusted. The dynamic coupling error between the material center temperature and the heat exchange jacket wall temperature is maintained within ±0.2℃.

[0048] Example 6: In a production preparation scenario adapting to the differences in oil absorption rates of chili pepper particles, the logic processing unit executes the angular momentum echo excitation intensity calibration procedure, controlling the torque load execution module to execute gradient deceleration pulses in the initial uniform speed segment, increasing the deceleration amplitude from 10% of the rated speed in 5% increments to 30%. The torque sensor monitors the cavitation characteristic frequency in the feedback signal. When the deceleration pulse width is 0.5s and the deceleration amplitude is 20%, the angular momentum echo component... The signal-to-noise ratio reaches 18dB. The logic processing unit locks the pulse intensity as the detection reference and, according to the formula... Calculate the uniformity correlation coefficient of the material in its initial state. ,in, The uniformity correlation coefficient is a dimensionless parameter. The angular momentum echo component is expressed in N·m·s. The starting time of the pulse deceleration action is in seconds, Δt is the monitoring duration, set to 2 seconds, and η is the thermal resistance coupling factor.

[0049] When the system encounters a situation where changing the chili variety alters the solid-liquid two-phase rheological properties, the logic processing unit initiates an online verification procedure for the phase interference constant β. A standard air flow rate of 0.15 MPa with a pulse width of 1 second is input through the bottom air inlet to simulate the bubble density distribution during peak microbial gas production. Simultaneously, the high-frequency fluctuation component of the instantaneous current from the stirring unit is collected. To calculate the characteristic entropy of fluctuation The system is based on the formula Calculate the phase disturbance constant β, where β is the phase disturbance constant and δ is the viscosity correction coefficient. The system calculates the β value as the fluctuation characteristic entropy and compares the deviation between the measured torque drop value and the loss compensation curve. If the deviation rate is within ±2.5%, the system stores the calibrated β value in the feature matrix of the logic processing unit. This enables the control system to adapt to different raw material rheological sensitivities. Regarding the extraction mechanism of the fluctuation characteristic entropy Sg, the logic processing unit pre-detects the high-frequency fluctuation component of the instantaneous current. Within the steady-state, disturbance-free range, the upper and lower limit amplitude envelopes divide the dynamic range of the current fluctuation into 64 discrete amplitude quantization intervals. Then, within each monitoring cycle, the microprocessor counter accumulates the frequency of valid sampling points falling into a specific quantization interval within the current 500ms sliding time window (i.e., 500 high-frequency discrete sampling points). This frequency is then divided by the total number of samples within the window to obtain the corresponding probability density pi. This deterministic array division rule and frequency distribution statistics directly constitute the mathematical input set for the subsequent information entropy formula calculation, giving the abstract entropy concept a precise industrial computing basis.

[0050] The embodiments of this application have been described above with reference to the accompanying drawings. Unless otherwise specified, the embodiments and features in the embodiments of this application can be combined with each other. This application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit of this application and the scope of protection of this invention, and all of these forms are within the protection scope of this application.

Claims

1. A fermentation production control system for chili oil, characterized in that, include: Logic processing unit, disturbance frequency modulation module, and torque load execution module; The logic processing unit is electrically connected to both the disturbance frequency modulation module and the torque load execution module. The logic processing unit is used to execute the following logic: Step 101: Output a perturbation command to the perturbation frequency modulation module to control the torque load execution module to superimpose a perturbation frequency signal with an amplitude of 3.5% of the rated current and a frequency in the range of 12Hz to 15Hz at the current speed, and obtain the real-time response torque data of the torque load execution module. Step 102: Extract the dynamic phase difference between the real-time response torque data and the drive frequency signal based on the perturbation frequency signal; Step 103, the reference self-calibration step, uses the dynamic phase difference to calculate the effective damping component caused by the controlled viscous medium contained in the torque load execution module, and corrects the physical baseline offset error based on the deviation between the effective damping component and the preset mechanical loss model, thus completing the zero-point self-calibration of the adjustment loop. A preset mechanical loss model is used to store the mapping relationship between the speed and the reference torque of the torque load execution module under no-load conditions; Step 104: Under the reference of zero-point self-calibration, determine the fermentation viscosity state based on the corrected effective damping component, and output control commands for adjusting the output torque of the torque load execution module or adjusting the heat exchange power.

2. The chili oil fermentation production control system according to claim 1, characterized in that, When the logic processing unit executes step 102, it performs a fast Fourier transform on the real-time response torque data to extract the phase offset angle at the frequency point of the perturbation frequency signal; and identifies the degree of nonlinear distortion of the phase offset angle relative to the perturbation frequency command to quantify the physical friction attenuation of the transmission chain of the torque load execution module; wherein, the physical friction attenuation is used as a dynamic correction parameter of the preset mechanical loss model to update the physical baseline offset error in step 103.

3. The chili oil fermentation production control system according to claim 1, characterized in that, The logic processing unit is also used to execute the angular momentum echo analysis logic: control the torque load execution module to perform pulse deceleration action and collect the damping attenuation curve of the controlled viscous medium under inertial residual; identify the uniformity of the internal flow field based on the damping attenuation curve, and extend the rotation switching cycle of the torque load execution module when the uniformity is lower than the preset threshold.

4. The chili oil fermentation production control system according to claim 1, characterized in that, The logic processing unit includes a shear sensitivity correction module, which is used to correct the rheological properties of the effective damping component to obtain the corrected fermentation viscosity. The calculation logic is as follows: ,in, The initial viscosity value is calculated based on the effective damping component, β is the shear thinning sensitivity factor of the controlled viscous medium, and Δω is the rotational speed fluctuation of the torque load execution module during the detection cycle; step 104 is based on the corrected fermentation viscosity. Determine the fermentation viscosity state.

5. The chili oil fermentation production control system according to claim 4, characterized in that, When the logic processing unit identifies the central thickening phenomenon inside the material based on the fermentation viscosity state, it outputs an increase command to control the torque load execution module to increase the instantaneous shear rate in order to break the phase accumulation inside the material.

6. The chili oil fermentation production control system according to claim 1, characterized in that, The system also includes a heat exchange regulation module, which is connected to the logic processing unit.

7. The chili oil fermentation production control system according to claim 1, characterized in that, The center frequency of the perturbation frequency signal avoids the mechanical resonance point of the transmission components inside the torque load execution module, so as to ensure that the signal-to-noise ratio of the real-time response torque data is not lower than the preset signal-to-noise ratio threshold.

8. The chili oil fermentation production control system according to claim 1, characterized in that, The logic processing unit performs a baseline self-calibration step once in each fermentation control cycle and stores the calibrated zero-point parameters in the real-time database.

9. A chili oil fermentation production control system according to claim 1, characterized in that, When the logic processing unit detects that the physical baseline offset error exceeds the safety limit deviation, it outputs a warning signal indicating that the signal strength attenuation of the sensing hardware has exceeded the limit.

10. A chili oil fermentation production control system according to claim 1, characterized in that, The system also includes a power module, which is electrically connected to the logic processing unit and the perturbation frequency modulation module.

Images