An intelligent control system for flour production process parameters

Abstract

The present application relates to the field of grain processing control, and discloses a flour production process parameter intelligent control system, comprising a multi-dimensional perception module, an operation control module and a collaborative execution module, the multi-dimensional perception module is used for obtaining physical basic data and defining physical monitoring boundaries, the operation control module converts the physical basic data into characteristic values representing material moisture gradient and dew point temperature, according to which the generated instruction bias of the vaporization amount generated by mechanical work is estimated, the bias instruction is generated in combination with the flow ratio deviation, and the bias instruction is directionally distributed according to the thermodynamic boundary condition, and the collaborative execution module receives the corresponding instruction to execute physical action. The present application fuses feedforward compensation and closed-loop feedback to compensate for process disturbance in advance, adopts a differentiated adjustment strategy to prevent high-temperature and high-humidity gas from being retained and dewing, simultaneously quantifies the safety margin of the microenvironment by using the thermodynamic boundary condition, avoids the powder road blockage caused by the adjustment under the dewing risk, and ensures the safety of system operation.

Description

Technical Field

[0001] This invention relates to the field of grain processing control, specifically to an intelligent control system for flour production process parameters. Background Technology

[0002] The grinding process in flour production is a core element determining flour quality and efficiency. During the extrusion and shearing of materials in the bran mill, the moisture distribution of the wheat grains, the mechanical parameters of the equipment, and changes in the temperature and humidity of the operating environment all directly affect the final grinding effect. Currently, the control of flour production process parameters typically relies on a single closed-loop feedback adjustment mechanism, that is, adjusting the operating parameters of the mechanical actuators in reverse based on the final state of the output material in order to maintain the set production targets.

[0003] Existing control methods suffer from response lag, failing to anticipate the impact of compression and shear thermal stress on materials before changes in their state, and thus unable to compensate for process disturbances in advance. Furthermore, conventional detection methods struggle to accurately characterize the difference in moisture distribution between the wheat kernel bran and endosperm, and at the execution level, they are limited to single-dimensional adjustments of mechanical mechanisms, lacking coordinated mapping control of pneumatic exhaust and cooling water circuits. This leads to the retention of high-temperature, high-humidity gases when mechanical work generates heat and moisture. Moreover, when the air in the grinding chamber's microenvironment is insufficient to prevent condensation and moisture precipitation, existing systems lack thermodynamic boundary condition-based distribution mechanisms and safety protection logic. Forcibly adjusting mechanical parameters under condensation risk can cause blockages in the grinding chamber's clogging rollers or powder paths, and in adverse operating conditions where the situation cannot be mitigated, the systems cannot physically shut off the actuators to ensure operational safety.

[0004] In summary, conventional flour production control mechanisms lack quantitative assessment of the variation patterns of pre-process physical quantities and have limitations in multi-dimensional execution coordination and underlying logic protection under extreme conditions. Therefore, how to provide an intelligent control system for flour production process parameters that integrates feedforward compensation and multi-dimensional collaborative execution, and possesses micro-environment thermodynamic boundary protection logic, is a problem that needs to be solved in this field. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides an intelligent control system for flour production process parameters, which solves the problems of existing systems having delayed response due to single feedback, inability to predict and compensate for process disturbances in advance, and lack of multi-dimensional collaborative adjustment and underlying safety protection logic under condensation risk conditions.

[0006] To address the above problems, the present invention provides the following technical solution: This invention provides an intelligent control system for flour production process parameters, employing the following technical solution: A smart control system for flour production process parameters includes: The multi-dimensional sensing module is used to acquire basic physical data in the production section and define the physical monitoring boundaries of the intelligent control system for flour production process parameters. The operation and control module is used to receive the physical basic data acquired by the multi-dimensional sensing module as raw data, convert the raw data into feature values ​​characterizing the moisture gradient of the material and the dew point temperature, estimate the amount of vaporization generated by mechanical work based on the feature values ​​characterizing the moisture gradient of the material, generate an instruction bias based on the amount of vaporization, calculate the flow ratio deviation of the fluid mass based on the raw data, perform proportional-integral-differential operation of the main loop to generate a deviation instruction, generate thermodynamic boundary conditions based on the dew point temperature, and directionally allocate the deviation instruction based on the thermodynamic boundary conditions. A collaborative execution module is used to receive the instruction bias and the deviation instruction, and to perform physical actions according to the instruction bias and the deviation instruction; The multidimensional perception module, the computation control module, and the collaborative execution module are connected in sequence.

[0007] By adopting the above technical solution, the multi-dimensional sensing module acquires the physical data of the production section. Based on this, the calculation and control module transforms the physical data into characteristic values ​​representing the material's moisture gradient and the dew point temperature. This allows for the pre-estimation of the vaporization generated by mechanical work and the generation of the command bias. Simultaneously, it combines the flow ratio deviation in the closed-loop circuit to generate the deviation command. Finally, it relies on the thermodynamic boundary conditions to directionally allocate the deviation command. Therefore, a control effect integrating feedforward compensation and closed-loop control is achieved. The intelligent control system for flour production process parameters can predict the extrusion and shear thermal effects on the material before significant state changes occur. It compensates for process disturbances in advance through the command bias and dynamically adjusts the distribution direction of control commands according to the thermodynamic state. This avoids the lag in response of single feedback control and system instability caused by adjustments under abnormal operating conditions, improving the uniformity of flour texture and process stability in the flour production process.

[0008] Furthermore, the multi-dimensional sensing module includes an electromagnetic sensing unit arranged at the discharge end of the wheat silo, a fluid sensing unit arranged at the two outlets of the high-square flat screen, and an environmental sensing unit arranged inside the grinding chamber of the first-stage mill and in the pipelines of the production workshop. The physical data includes low-frequency AC impedance sequences and high-frequency AC impedance sequences obtained by the electromagnetic sensing unit, coarse powder mass flow rate and fine powder mass flow rate obtained by the fluid sensing unit, and grinding roller surface temperature, ambient temperature, and relative humidity obtained by the environmental sensing unit. The collaborative execution module includes a mechanical execution unit, a pneumatic execution unit, and a thermal control execution unit. The mechanical execution unit is used to adjust the grinding roller spacing and grinding roller speed difference of the first-stage mill. The pneumatic execution unit includes a variable frequency pneumatic regulating valve for adjusting the exhaust volume and air pressure of the grinding chamber suction pipe. The thermal control execution unit includes a proportional flow regulating valve for adjusting the cooling water inlet volume inside the grinding roller.

[0009] By adopting the above technical solution, a sensor monitoring network covering multiple dimensions of material dielectric properties, fluid dynamics, and thermodynamics was established, and matched with a multi-dimensional actuator covering mechanical mechanisms, pneumatic ventilation, and cooling water circuits. The low-frequency AC impedance sequence and the high-frequency AC impedance sequence reflect the difference in moisture distribution between the material's surface and interior, providing a physical input source for process prediction; at the same time, the collaborative execution architecture of mechanical, pneumatic, and thermal control dimensions overcomes the limitations of relying on adjusting the grinding roller spacing to control production, providing physical execution support for targeted intervention of subsequent commands.

[0010] Furthermore, the operation control module includes: a state reconstruction unit, which is used to convert the low-frequency AC impedance sequence, the high-frequency AC impedance sequence, the ambient temperature and the relative humidity into cortex toughness characteristic value, endosperm compressive strength characteristic value and the dew point temperature, wherein the cortex toughness characteristic value and the endosperm compressive strength characteristic value constitute the characteristic value characterizing the moisture gradient of the material; a pre-operation unit, which is used to estimate the amount of vaporization generated by mechanical work based on the cortex toughness characteristic value and the endosperm compressive strength characteristic value, and generate the instruction bias amount based on the amount of vaporization; and a closed-loop feedback unit, which is used to calculate the flow ratio deviation based on the mass flow rate of coarse powder and the mass flow rate of fine powder, perform proportional-integral-differential operation of the main loop to generate the deviation instruction, generate the thermodynamic boundary conditions based on the ambient temperature and the dew point temperature, and perform directional allocation of the deviation instruction based on the thermodynamic boundary conditions.

[0011] By adopting the above technical solution, a clear hierarchical logic is formed within the computational control module. The state reconstruction unit converts electrical signals and environmental parameters into digital parameters characterizing the physical properties of the materials; the pre-processing unit performs feedforward control, using physical property parameters to predict vaporization interference and generate initial compensation; the closed-loop feedback unit performs feedback control and safety allocation, calculating and converging control errors based on deviations in production results. This architecture balances control response speed and long-term operational accuracy.

[0012] Furthermore, the state reconstruction unit is specifically used to: perform a quotient operation between the low-frequency AC impedance sequence and the high-frequency AC impedance sequence to construct an impedance ratio sequence; use the impedance ratio sequence and the ambient temperature as input variables, and calculate the cortical toughness characteristic value and the endosperm compressive strength characteristic value through a polynomial reconstruction equation; and mathematically couple the ambient temperature and the relative humidity to calculate the dew point temperature.

[0013] By employing the above technical solution, background errors such as absolute moisture content are eliminated through the quotient operation of the low-frequency AC impedance sequence and the high-frequency AC impedance sequence, enabling the constructed impedance ratio sequence to accurately characterize the relative distribution differences between the moisture content in the wheat grain cortex and the endosperm. Reconstruction based on ambient temperature corrects the interference of temperature changes on the material's mechanical properties, reflecting the mechanical characteristics of the material under different moisture states, and providing a data source for subsequent calculations of the cortex toughness characteristic value and the endosperm compressive strength characteristic value.

[0014] Furthermore, the pre-processing unit is specifically used to: calculate the temperature difference between the surface temperature of the grinding roller and the ambient temperature when estimating the amount of vaporization generated by mechanical work; perform linear weighted coupling calculation on the skin toughness characteristic value, the endosperm compressive strength characteristic value and the temperature difference to generate a phase change vaporization prediction index; and use the phase change vaporization prediction index as the amount of vaporization.

[0015] By employing the above technical solution, the mechanical properties of the material are combined with the thermodynamic state of the production equipment to quantitatively evaluate the energy conversion during the grinding process. The skin toughness characteristic value and the endosperm compressive strength characteristic value correspond to the amount of mechanical energy required for grinding, while the temperature difference reflects the current heat dissipation conditions of the equipment. The phase change vaporization prediction index generated by the coupling of these two factors reflects the intensity of gas generation due to the conversion of mechanical energy into heat energy, resulting in the evaporation of moisture from the material, thus forming the basis for feedforward compensation.

[0016] Furthermore, when generating the instruction bias based on the vaporization amount, the pre-processing unit specifically performs the following: using the phase change vaporization prediction index, it calculates the roller spacing bias corresponding to the extrusion pressure and the roller speed difference bias corresponding to the shear force through algebraic scaling operations; using the phase change vaporization prediction index, it generates a pre-opening instruction for the variable frequency pneumatic regulating valve through an algebraic mapping equation with quadratic amplification characteristics; using the phase change vaporization prediction index, it generates an inverse proportional opening instruction for the proportional flow regulating valve through a mapping equation based on an inverse proportional algebraic structure; the instruction bias includes the roller spacing bias, the roller speed difference bias, the pre-opening instruction for the variable frequency pneumatic regulating valve, and the inverse proportional opening instruction for the proportional flow regulating valve.

[0017] By adopting the above technical solutions, a nonlinear mapping strategy is employed for different actuators to address process interference caused by vaporization. For the variable frequency pneumatic regulating valve, a quadratic amplification characteristic regulation strategy is used to increase the suction intensity when the anticipated increase in vaporization is achieved, preventing the retention and condensation of high-temperature and high-humidity gases. For the proportional flow regulating valve, inverse proportional regulation is used to enhance the heat dissipation capacity of the grinding rollers. Scaling compensation is applied to the mechanical actuator to offset the impact of material property changes in the initial stage of grinding work, maintaining stable production.

[0018] Furthermore, the closed-loop feedback unit, when calculating the flow ratio deviation and generating the deviation command, specifically performs the following: algebraically quotient coupling of the fine powder mass flow rate and the coarse powder mass flow rate to calculate the actual mass flow rate ratio; algebraically subtracting the preset target mass flow rate ratio from the actual mass flow rate ratio to calculate the flow ratio deviation; using the flow ratio deviation as the excitation source for closed-loop control to perform discrete position proportional-integral-differential operations to generate a comprehensive feedback control command; and using the comprehensive feedback control command as the deviation command.

[0019] By adopting the above technical solution and using the fluid mass flow rate ratio as the main control variable, the actual scraping rate and powder output efficiency of the grinding process are reflected. The discrete position proportional-integral-differential algorithm is used to process the flow rate ratio deviation, which can eliminate steady-state errors and output smooth integrated feedback control commands. This avoids high-frequency oscillations in the actuator due to natural fluctuations in the material conveying process, ensuring the continuity of process control.

[0020] Furthermore, the closed-loop feedback unit, when generating the thermodynamic boundary conditions, is specifically used to: calculate the algebraic difference between the ambient temperature and the dew point temperature to obtain a thermodynamic temperature difference margin; compare the thermodynamic temperature difference margin with a preset safe temperature difference critical threshold; when the thermodynamic temperature difference margin is less than or equal to the safe temperature difference critical threshold, set a high humidity danger flag in the status register; when the thermodynamic temperature difference margin is greater than the safe temperature difference critical threshold, reset the high humidity danger flag in the status register; the state of the high humidity danger flag constitutes the thermodynamic boundary conditions.

[0021] By adopting the above technical solution, the safe margin between the air inside the grinding chamber and the condensation of moisture is quantified. The smaller the thermodynamic temperature difference margin value, the closer the microenvironment is to the dew point temperature. By setting a critical threshold and driving the high humidity danger flag in the status register to be set and reset, the intelligent control system for flour production process parameters can obtain the status judgment result before gas condensation, providing a logical boundary for subsequent adjustment of the control strategy.

[0022] Furthermore, when the closed-loop feedback unit allocates the deviation command according to the thermodynamic boundary conditions, it specifically performs the following: when the high humidity hazard flag is in a reset state, the closed-loop feedback unit allocates the deviation command to the mechanical execution unit and calculates the deviation command by superimposing it with the command bias amount corresponding to the mechanical execution unit; when the high humidity hazard flag is in a set state, the closed-loop feedback unit freezes the allocation of the deviation command to the mechanical execution unit and allocates the deviation command to the pneumatic execution unit and the thermal control execution unit, and calculates the deviation command by superimposing it with the command bias amounts corresponding to the pneumatic execution unit and the thermal control execution unit.

[0023] By adopting the above technical solution, the intelligent control system for flour production process parameters possesses the ability to switch control paths based on environmental safety conditions. Under normal thermodynamic conditions, the deviation commands are primarily used to adjust the mechanical execution units to ensure flour output indicators. However, when facing the risk of high humidity and condensation, the load increase action of the mechanical execution units is frozen, and the deviation commands are instead directed to the pneumatic execution units and the thermal control execution units. This allocation mechanism avoids the problem of grinding roller clogging or powder path blockage caused by adjusting mechanical parameters under critical condensation conditions.

[0024] Furthermore, the closed-loop feedback unit is also equipped with a safety timer; when the high humidity hazard flag is in the set state, the closed-loop feedback unit starts the safety timer to start timing; when the cumulative timing duration of the safety timer exceeds a preset limit safety threshold, the closed-loop feedback unit generates a system shutdown command and sends the system shutdown command to the collaborative execution module, and the collaborative execution module stops executing the physical action according to the system shutdown command.

[0025] By adopting the above technical solution, a blocking mechanism for harsh working conditions is introduced. When the adjustment of airflow and cooling water fails to change the high humidity condensation trend of the microenvironment within a specified time, the safety timer will trigger protection. The collaborative execution module stops the physical execution of the current section, preventing production accidents from occurring at the physical level and ensuring the overall operational safety of the system.

[0026] This invention provides an intelligent control system for flour production process parameters. It has the following beneficial effects: 1. This invention acquires physical data through a multi-dimensional sensing module, and the calculation and control module transforms the physical data into characteristic values ​​representing the moisture gradient of the material and the dew point temperature. Based on this, the amount of vaporization generated by mechanical work is estimated to generate a command bias, and a deviation command is generated by combining the flow ratio deviation. Then, the deviation command is directionally allocated according to the thermodynamic boundary conditions. This method realizes the integration of feedforward compensation and closed-loop feedback, which can predict the thermal impact on the material before the material state changes and compensate for process disturbances in advance. It avoids the response lag of single feedback control and improves the process stability of the control system.

[0027] 2. This invention constructs an impedance ratio sequence using a state reconstruction unit with low-frequency and high-frequency AC impedance sequences, calculates cortical toughness characteristic values ​​and endosperm compressive strength characteristic values ​​based on ambient temperature, and generates instruction bias values ​​containing multi-dimensional execution dimensions using a phase change vaporization prediction index by a pre-processing unit. This method accurately characterizes the differences in moisture distribution in different parts of the wheat grain and adopts differentiated algebraic mapping adjustment strategies for pneumatic execution units, thermal control execution units, and mechanical execution units, which not only offsets the impact of changes in material properties but also prevents high-temperature and high-humidity gas stagnation and condensation.

[0028] 3. This invention uses a closed-loop feedback unit to construct thermodynamic boundary conditions by comparing the thermodynamic temperature difference margin with the critical threshold of safe temperature difference. When the high humidity danger flag is in the set state, the system freezes the distribution of deviation instructions to the mechanical execution unit and starts the safety timer. When the accumulated time exceeds the limit safety threshold, the system controls the collaborative execution module to stop performing physical actions. This method quantifies the safe margin of the air in the microenvironment from condensation and precipitation, avoids the problem of powder path blockage caused by forcibly adjusting mechanical parameters under the risk of condensation, and ensures operational safety from a physical level when severe working conditions cannot be alleviated. Attached Figure Description

[0029] Figure 1 This is a diagram illustrating the architecture of an intelligent control system for flour production process parameters according to an embodiment of the present invention. Figure 2 This is a flowchart illustrating the workflow of an intelligent control system for flour production process parameters according to an embodiment of the present invention. Figure 3 This is a comparison diagram of the thermodynamic temperature difference margin of one embodiment of the present invention; Figure 4 This is a comparison diagram of the underlying device instruction output actions according to an embodiment of the present invention; Figure 5 This is a comparison chart of the actual mass flow rate ratio under different strategies according to an embodiment of the present invention. Detailed Implementation

[0030] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0031] See attached document Figure 1 This invention provides an intelligent control system for flour production process parameters, comprising a multi-dimensional sensing module, a computational control module, and a collaborative execution module. The multi-dimensional sensing module, computational control module, and collaborative execution module are connected sequentially to realize the cross-correlation and dynamic allocation of physical parameters during flour processing.

[0032] The multi-dimensional sensing module is used to acquire basic physical data in the production section and define the physical monitoring boundaries of the system. The multi-dimensional sensing module includes electromagnetic sensing units, fluid sensing units, and environmental sensing units.

[0033] The hardware carrier of the electromagnetic sensing unit is a ring electrode array. The ring electrode array is arranged at the discharge end of the wheat storage bin. The electromagnetic sensing unit is used to acquire basic AC impedance data of the raw wheat material at different frequency bands.

[0034] The hardware carrier of the fluid sensing unit is a microwave solid mass flow meter. The microwave solid mass flow meter is arranged at both outlets of the high-square flat screen. The fluid sensing unit is used to obtain the real-time mass flow rate of the product screened by the high-square flat screen.

[0035] The environmental sensing unit's hardware consists of an infrared temperature sensor and a temperature and humidity transmitter. The infrared temperature sensor is located inside the grinding chamber of the first grinding mill. The temperature and humidity transmitter is located within the pipelines of the production workshop. The environmental sensing unit is used to acquire the solid-phase interface temperature and the ambient temperature and humidity of the gas phase environment in the first grinding mill.

[0036] The arithmetic control module is deployed inside the programmable logic controller (PLC). It is used to execute algebraic operations in the control logic program. The arithmetic control module includes a state reconstruction unit, a pre-processing unit, and a closed-loop feedback unit.

[0037] The state reconstruction unit is communicatively connected to the multidimensional sensing module. The state reconstruction unit receives the raw data acquired by the multidimensional sensing module and transforms it into characteristic values ​​representing the material's moisture gradient and dew point temperature through algebraic operations.

[0038] The pre-processing unit is connected to the state reconstruction unit. The pre-processing unit is used to estimate the amount of vaporization generated by mechanical work based on the characteristic values ​​output by the state reconstruction unit, and to generate the pre-processing instruction bias based on the amount of vaporization.

[0039] The closed-loop feedback unit is communicatively connected to the multi-dimensional sensing module. The closed-loop feedback unit is used to calculate the flow rate ratio deviation of the fluid mass, perform proportional-integral-differential (PID) calculations in the main loop, and directionally distribute deviation commands based on thermodynamic boundary conditions.

[0040] The collaborative execution module is connected to both the pre-processing unit and the closed-loop feedback unit in the arithmetic control module. It receives the instruction bias and deviation commands from the arithmetic control module and executes the corresponding physical actions. The collaborative execution module includes a mechanical execution unit, a pneumatic execution unit, and a thermal control execution unit.

[0041] The hardware of the mechanical actuator includes a position servo motor and a frequency converter. The position servo motor receives control commands and adjusts the grinding roller spacing of the first grinding mill. The frequency converter receives control commands and adjusts the grinding roller speed difference of the first grinding mill.

[0042] The hardware of the pneumatic actuator includes a variable frequency pneumatic control valve. This valve is located within the suction duct of the grinding chamber of the first grinding mill. The pneumatic actuator is used to regulate the exhaust volume and air pressure of the grinding chamber suction duct.

[0043] The hardware of the thermal control actuator includes a proportional flow regulating valve. This valve is located in the cooling water circulation line inside the grinding rollers of the first grinding mill. The thermal control actuator is used to adjust the flow rate of the cooling water inside the grinding rollers to change the heat conduction state.

[0044] See attached document Figure 2 This invention provides an intelligent control method for flour production process parameters, comprising the following steps: S10. When the raw wheat material flows from the discharge end of the wheat storage silo, the electromagnetic sensing unit in the multi-dimensional sensing module acquires the basic AC impedance data of the material according to the preset scanning sequence. The state reconstruction unit in the calculation control module receives the basic AC impedance data and converts it into a feature value that characterizes the moisture gradient, thus completing the pre-sensing and initial state parameter database construction. S20. As the physical conveying of materials generates a time delay and reaches the first grinding mill, the pre-processing unit in the calculation control module estimates the vaporization amount based on the characteristic value and generates a reference bias instruction. The calculation control module synchronously sends the reference bias instruction to the mechanical execution unit, pneumatic execution unit and thermal control execution unit in the collaborative execution module for feedforward collaborative action execution. S30. After the material enters the high-square flat screen for screening, the fluid sensing unit in the multi-dimensional sensing module acquires the real-time mass flow rate data of the product and returns it to the calculation and control module. The closed-loop feedback unit in the calculation and control module calculates the flow ratio deviation and the current temperature difference boundary judgment condition based on the data acquired by the environmental sensing unit. The closed-loop feedback unit selectively drives the corresponding execution unit in the collaborative execution module to perform physical action correction based on the temperature difference boundary judgment condition, thereby completing the closed-loop compensation and instruction orientation allocation.

[0045] To further illustrate the underlying data processing mechanism and internal cross-domain allocation logic of this invention, the calculation rules and specific implementation details of each step in the above workflow will be elaborated below.

[0046] In flour production, the moisture penetration state of raw wheat after resting in the wheat conditioning silo directly determines the setting of subsequent grinding and sieving process parameters. To accurately quantify the internal moisture distribution of the material, the computational control module needs to extract features based on the underlying physical data acquired by the multi-dimensional sensing module. S10 acquires electromagnetic response data of raw wheat material at different AC frequency bands to provide a basic data source for the state reconstruction unit, completing the pre-sensing and initial state parameter database construction.

[0047] S101, the electromagnetic sensing unit in the multi-dimensional sensing module alternately injects high and low frequency AC excitation signals according to the preset scanning sequence and obtains the original impedance sequence.

[0048] The lower-level hardware carrier of the electromagnetic sensing unit is a ring electrode array. The ring electrode array is fitted onto the inner wall cross-section of the wheat silo's discharge end, ensuring that the raw wheat material flows through the detection space formed by the ring electrode array during its descent under gravity. The signal generation circuit inside the arithmetic control module applies an AC excitation signal to the ring electrode array.

[0049] Different frequencies of alternating electromagnetic fields exhibit significant differences in their polarization penetration ability towards free water on the surface and bound water within wheat grains. The operational control module employs a time-division multiplexing mechanism to alternately inject two fixed-frequency alternating current excitation signals into the ring electrode array. These two fixed frequencies include a low-frequency test frequency and a high-frequency test frequency.

[0050] The computation control module continuously advances the current sampling time according to a fixed sampling period. Within the period corresponding to each current sampling time, the computation control module divides the time axis into a first test time window and a second test time window.

[0051] During the first test time window, the computational control module injects a low-frequency AC excitation signal into the ring electrode array. The multi-dimensional sensing module simultaneously acquires the voltage and current feedback values ​​flowing through the raw wheat material, and generates a low-frequency AC impedance sequence corresponding to the current sampling time after analog-to-digital conversion.

[0052] During the second test time window, the operational control module injects a high-frequency AC excitation signal into the ring electrode array. The multi-dimensional sensing module acquires the voltage and current feedback values ​​under the high-frequency response based on the same acquisition rules, generating a high-frequency AC impedance sequence corresponding to the current sampling moment.

[0053] S102, the state reconstruction unit in the operation control module receives the low-frequency AC impedance sequence and the high-frequency AC impedance sequence, and generates the impedance ratio sequence through algebraic operations.

[0054] The state reconstruction unit, through its internally configured arithmetic logic unit, performs a quotient operation between the low-frequency AC impedance sequence corresponding to each current sampling moment and the high-frequency AC impedance sequence corresponding to the same current sampling moment, constructing an impedance ratio sequence reflecting the material characteristics. The specific algebraic formula is as follows: ; in, This is the current sampling time; The current sampling time The corresponding impedance ratio sequence; The current sampling time The corresponding low-frequency AC impedance sequence; The current sampling time The corresponding high-frequency AC impedance sequence.

[0055] Alternating current signals at different frequency bands exhibit specific dielectric polarization effects when penetrating raw wheat material. At low frequency test frequencies, the AC excitation signal mainly propagates along the surface of the raw wheat material, and its impedance value is directly affected by the free water content in the outer layer. At high frequency test frequencies, the AC excitation signal penetrates the outer layer of the raw wheat material and enters the internal region, and its impedance value is directly affected by the bound water content in the internal endosperm region. The impedance ratio sequence generated by the state reconstruction unit using quotient operations can eliminate common-mode data interference caused by changes in the absolute volume and porosity of the raw wheat material, allowing this impedance ratio sequence to directly map and quantify the moisture permeation gradient of the internal and external moisture distribution of the raw wheat material mathematically.

[0056] S103, the state reconstruction unit in the operation control module uses the impedance ratio sequence and the ambient temperature obtained by the multidimensional sensing module to calculate the cortical toughness characteristic value and the endosperm compressive strength characteristic value through the polynomial reconstruction equation.

[0057] The dielectric constant and physical rheological state of moisture within raw wheat material are influenced by the gas-phase thermodynamic environment. To eliminate the interference of temperature drift on moisture state assessment, the floating-point arithmetic processor within the state reconstruction unit uses the impedance ratio sequence corresponding to the current sampling moment, along with the ambient temperature corresponding to the current sampling moment synchronously acquired by the environmental sensing unit in the multi-dimensional sensing module, as input variables.

[0058] The toughness of the cortex is directly related to the degree of free water in the outer layer of the raw wheat material, and this degree of free water has a positive linear mapping relationship with the impedance ratio. The compressive strength of the endosperm is mainly related to the tightness of the binding between bound water and starch granules within the raw wheat material, and this tightness of binding has an inversely proportional nonlinear mapping relationship with the impedance ratio. Based on the above physical mapping relationships, the state reconstruction unit executes the following polynomial reconstruction equation: ; ; in, This is the current sampling time; The current sampling time The corresponding cortical toughness characteristic value; The current sampling time The corresponding endosperm compressive strength characteristic value; The current sampling time The corresponding impedance ratio sequence; The current sampling time The corresponding ambient temperature; , , These are the constant coefficients in the first constant coefficient matrix that are pre-calibrated for a specific wheat variety; , , These are the constant coefficients in the second constant coefficient matrix that are pre-calibrated for a specific wheat variety.

[0059] The state reconstruction unit completes the transformation of multidimensional physical quantities from the underlying detection signal to the engineering control characteristic value through the above algebraic operations. It stores the cortical toughness characteristic value and the endosperm compressive strength characteristic value that characterize the moisture gradient into the internal register of the programmable logic controller as the basis for subsequent pre-bias allocation.

[0060] S104. The environmental sensing unit in the multi-dimensional sensing module obtains the ambient temperature and relative humidity corresponding to the current sampling time. The state reconstruction unit in the calculation and control module combines the ambient temperature and relative humidity to calculate the dew point temperature corresponding to the current production condition in real time.

[0061] The temperature and humidity transmitter in the environmental sensing unit transmits the collected analog signals to the state reconstruction unit after analog-to-digital conversion. The critical thermodynamic threshold for the phase change and condensation of gaseous water molecules in the flour production workshop and exhaust duct changes dynamically with the environmental conditions. To accurately obtain this critical thermodynamic threshold, the state reconstruction unit calls logarithmic and fractional operation instructions through its internal mathematical function library. Based on Magnus's empirical principle, it mathematically couples the ambient temperature and relative humidity at the current sampling time to calculate the dew point temperature at the current sampling time.

[0062] The dew point temperature calculated by the state reconstruction unit defines the precise temperature at which water vapor in the gaseous environment reaches saturation. The state reconstruction unit stores the calculated dew point temperature in the global data block of the programmable logic controller. This dew point temperature will serve as a reference parameter for subsequent steps in calculating temperature difference boundaries and determining whether there is a risk of condensation adhesion in the flow field.

[0063] The AC impedance measurement bridge or conversion chip application, the moving average filtering and noise reduction of digital signals, the offline fitting and extraction of constant coefficients based on the standard oven method and the least squares method, and the underlying physical sensing and digital conversion mechanism of commercially available industrial-grade temperature and humidity probes involved in the above S10 steps are all well-known technologies in the field. Those skilled in the art can implement them using conventional means, and will not be elaborated here.

[0064] After the raw wheat material completes state characteristic extraction and parameter database construction in the wheat replenishment warehouse, it is transported to the first tanning mill via workshop pipelines. During the delay caused by this physical transport, the internal moisture distribution characteristics of the material have solidified. To address the moisture phase change vaporization phenomenon caused by the large amount of frictional heat generated by mechanical crushing and work upon entering the grinding chamber of the first tanning mill, the computational control module makes a prediction and intervenes in advance. S20 details the internal mapping mechanism by which the pre-processing unit uses the acquired state characteristic values ​​to estimate the amount of phase change vaporization and synchronously issues bias commands to the execution units of the three different physical domains: mechanical, pneumatic, and thermal control.

[0065] S201, the pre-processing unit in the computation control module uses the cortical toughness characteristic value, the endosperm compressive strength characteristic value, and the real-time temperature difference between the solid and gas phases to construct a multivariate equation logic for the phase change vaporization prediction index.

[0066] The environmental sensing unit in the multi-dimensional sensing module uses an infrared temperature sensor arranged inside the grinding chamber of the first grinding mill to obtain the surface temperature of the grinding rollers in real time at the current sampling moment. Since there is a fixed transmission time for the material to physically flow from the outlet of the wheat silo to the first grinding mill, the pre-processing unit performs calculations at the current sampling moment, and the extracted material physical characteristic data matches the material that actually arrives at the grinding chamber.

[0067] The pre-processing unit is equipped with a time compensation queue. Based on the material conveying delay time calibrated by the system, it retrieves the cortical toughness characteristic value and endosperm compressive strength characteristic value corresponding to the delayed sampling time obtained at the discharge end of the wheat silo for this batch of materials from the internal register.

[0068] As the material enters the grinding chamber, mechanical scraping and compression cause internal moisture to absorb frictional heat and convert into water vapor. The surface toughness characteristic value positively determines the ease with which surface moisture is peeled off and released; the endosperm compressive strength characteristic value positively determines the mechanical energy consumed and the frictional heat generated during particle crushing; the temperature difference between the grinding roller surface temperature and the ambient temperature constitutes the driving force for heat exchange in the thermodynamic system. The pre-processing unit substitutes the above variables into the internal multivariate equations for linear weighted coupling calculations.

[0069] The multivariate equation formula for the phase change vaporization prediction exponent, specifically executed by the pre-processing unit, is as follows: ; in, This is the current sampling time; For material conveying delay time; For delayed sampling time; The current sampling time The corresponding phase change vaporization prediction index; Delayed sampling time The corresponding cortical toughness characteristic value; Delayed sampling time The corresponding endosperm compressive strength characteristic value; The current sampling time The corresponding surface temperature of the grinding roller; The current sampling time The corresponding ambient temperature; These are the first weighting coefficients of the multivariate prediction equation; These are the second weighting coefficients of the multivariate prediction equation; The third weighting coefficient is used in the multivariate prediction equation.

[0070] The phase change vaporization prediction index calculated by the pre-processing unit mathematically and quantitatively characterizes the expected mass scale of water vapor released into the gas phase space of the grinding chamber at the moment of crushing by the material currently entering the first grinding mill. The phase change vaporization prediction index is written into the memory address of the control system and serves as the benchmark for the pre-processing unit to issue feedforward bias instructions for each physical domain to the lower-level collaborative execution module.

[0071] S202, the pre-processing unit in the operation and control module maps the phase change vaporization prediction index to the mechanical execution unit in the collaborative execution module to generate the position offset of the position servo motor and the speed ratio offset of the frequency converter.

[0072] A high phase change vaporization prediction index indicates that excessive frictional heat and moisture evaporation are expected to occur when the raw wheat material is scraped and crushed in the first hull mill. To suppress the increase in heat from the source physical and mechanical level, the pre-processing unit reduces the normal extrusion force and tangential shear force applied to the material by the grinding rollers in advance. The logic program inside the pre-processing unit uses the phase change vaporization prediction index as the independent variable and calculates the grinding roller spacing offset corresponding to the extrusion force and the grinding roller speed difference offset corresponding to the shear force through algebraic scaling operations.

[0073] The calculation formula for the feedforward mapping of the pre-processing unit for the specific execution of the mechanical execution unit is as follows: ; ; in, This is the current sampling time; The current sampling time The corresponding grinding roller spacing offset; This is the spacing adjustment ratio coefficient, whose value is preset to a positive value by the system. The current sampling time The corresponding phase change vaporization prediction index; The current sampling time The corresponding roller speed difference offset; This is the speed difference adjustment ratio coefficient, and its value is preset to a negative value by the system.

[0074] Based on the above mathematical mapping relationship, when the estimated degree of vaporization intensifies, the roller spacing offset calculated by the pre-processing unit increases in the positive direction, indicating that the gap between the fast and slow rollers of the first grinding mill needs to be finely adjusted to reduce mechanical normal extrusion; the roller speed difference offset calculated synchronously decreases in the opposite direction, indicating that the relative linear velocity difference between the fast and slow rollers needs to be appropriately reduced to weaken the tangential scraping and shearing work.

[0075] The pre-processing unit encapsulates the calculated roller spacing offset and roller speed difference offset into a low-level communication message. The processing control module uses an industrial Ethernet bus to send the communication message to the mechanical execution unit in the collaborative execution module. The position servo motor in the mechanical execution unit receives the roller spacing offset command and drives the mechanical eccentric shaft mechanism to fine-tune the physical spacing; the frequency converter in the mechanical execution unit receives the roller speed difference offset command and adjusts the output frequency to change the speed of the drive motor.

[0076] S203, the pre-processing unit in the operation and control module maps the phase change vaporization prediction index to the pneumatic execution unit in the collaborative execution module, and generates the pre-opening command of the variable frequency pneumatic regulating valve through an algebraic amplification algorithm.

[0077] During the flour milling process, even if the mechanical actuator reduces the work done by fine-tuning, the internal moisture of the raw wheat material will still be released into the gas phase space of the grinding chamber of the first-stage mill when it breaks down. To prevent high-concentration water vapor from condensing and adhering to the inner wall of the grinding chamber or the suction pipe, the pre-processing unit uses the phase change vaporization prediction index as a feedforward input to increase the exhaust suction force of the pneumatic system in advance. Considering the nonlinear expansion of water vapor volume when heated, the arithmetic logic unit inside the pre-processing unit is configured with an algebraic mapping equation with quadratic amplification characteristics for the pneumatic actuator to ensure that the incremental response speed of the pneumatic exhaust volume can cover the increase in water vapor volume.

[0078] The algebraic amplification calculation formulas for the pneumatic actuators executed by the pre-processing unit are as follows: ; in, This is the current sampling time; The current sampling time The corresponding variable frequency pneumatic control valve pre-opening command; This is the aerodynamic amplification gain coefficient, the value of which is pre-calibrated to a fixed positive value by the system based on the cross-sectional area of ​​the pipeline and the power of the fan. The current sampling time The corresponding phase change vaporization prediction index.

[0079] Based on the aforementioned mathematical formula, when the phase change vaporization prediction index increases, the pre-opening command of the variable frequency pneumatic regulating valve will amplify significantly according to a parabolic law. The calculation and control module sends the generated pre-opening command of the variable frequency pneumatic regulating valve to the pneumatic execution unit in the collaborative execution module. The variable frequency pneumatic regulating valve in the pneumatic execution unit drives the valve core displacement according to this bias command, expanding the flow cross-sectional area inside the grinding chamber suction pipe of the first grinding mill. The exhaust volume of the grinding chamber suction pipe then increases nonlinearly, forming forced convection, accelerating the extraction of the mixed gas carrying high humidity water vapor inside the grinding chamber, and disrupting the physical conditions for water vapor to reach saturation at the aerodynamic level.

[0080] S204. The pre-processing unit in the operation and control module maps the phase change vaporization prediction index to the thermal control execution unit in the collaborative execution module, generating the inverse proportional opening command of the proportional flow regulating valve.

[0081] During the initial peeling of raw wheat in the first grinding mill, the heat generated by mechanical friction is conducted to the material through the surface of the grinding rollers, becoming the main heat source driving the vaporization of moisture inside the material. To suppress moisture evaporation from the heat conduction physical domain, the pre-processing unit uses the phase change vaporization prediction index to adjust the flow rate of cooling water inside the grinding rollers in advance. When the phase change vaporization prediction index increases, it means that the expected frictional heat will increase significantly, and the system needs to provide a larger cooling water flow rate to enhance heat exchange. To achieve a nonlinear and rapid response to the heat conduction state, the pre-processing unit is configured with a mapping equation based on an inverse proportional algebraic structure, so that when the phase change vaporization prediction index increases, the inverse proportional term decays, thus making the final calculated regulating valve opening exhibit an amplified physical effect.

[0082] The formula for calculating the inverse proportional opening degree executed by the pre-processing unit for the thermal control execution unit is as follows: ; in, This is the current sampling time; The current sampling time The corresponding proportional flow control valve inverse proportional opening command; The thermal control basic opening constant is preset for the system; The thermal control inverse proportional adjustment coefficient is preset for the system; The current sampling time The corresponding phase change vaporization prediction index; To prevent positive real constants with a denominator of zero.

[0083] The computational control module converts the generated proportional flow regulating valve inverse proportional opening command into a standard current signal via the analog output channel and sends it to the thermal control execution unit in the collaborative execution module. The proportional flow regulating valve in the thermal control execution unit receives this standard current signal and proportionally drives the valve core displacement, expanding the flow cross-sectional area of ​​the cooling water circulation pipe inside the first grinding mill roller. As the cooling water inflow increases, the convective heat transfer coefficient inside the grinding roller rises, maintaining the solid-phase interface temperature within a safe thermodynamic boundary, thereby cutting off the energy source for the continuous phase change of moisture at the thermal conduction level.

[0084] The underlying signal conversion of the commercial non-contact infrared temperature measurement probe, the underlying electrical drive control of the standard vector frequency converter and AC servo system, the closed-loop control of the pneumatic diaphragm actuator and its electrical intelligent positioner, and the electromagnetic drive and fluid throttling control of the industrial-grade electromagnetic proportional water valve involved in the above S20 steps are all well-known technologies in the field. Those skilled in the art can directly implement them using commercially available standard parts and conventional methods, and will not be elaborated here.

[0085] After the raw wheat material is ground and crushed in the first hull mill, the mixture enters the high-square flat screen for screening via a physical conveying pipeline. The pre-positioned feedforward bias control can intervene early and effectively suppress transient moisture phase changes and heat increases. However, due to fluctuations in the physical properties of the raw wheat material and the inherent hysteresis of the mechanical actuators, the system will generate steady-state errors during long-term operation. To eliminate steady-state errors and prevent physical action conflicts between multiple actuators when correcting deviations, the computational control module introduces a closed-loop feedback and anti-conflict mechanism. S30 details the complete physical correction process of the closed-loop feedback unit performing main loop deviation calculations based on the screening product data from the high-square flat screen, and implementing directional allocation of the comprehensive feedback control commands in conjunction with thermodynamic boundary conditions.

[0086] S301, the fluid sensing unit in the multi-dimensional sensing module acquires real-time mass flow data of the two outlets of the high-square flat screen, and the closed-loop feedback unit in the calculation control module continuously calculates the actual mass flow ratio corresponding to the current production state based on the two real-time mass flow data.

[0087] The lower-level hardware carrier of the fluid sensing unit in the multi-dimensional sensing module is a microwave solid mass flow meter. The microwave solid mass flow meter is vertically attached to the outer wall of the coarse powder discharge pipe and the fine powder discharge pipe at the bottom of the high-square flat screen, respectively. The fluid sensing unit emits microwave signals into the discharge pipe and receives the echo signals reflected by the falling material, continuously acquiring the mass of solid powder flowing in the discharge pipe.

[0088] The fluid sensing unit converts the acquired fluid quality data into standard digital signals and transmits them to the arithmetic control module. The closed-loop feedback unit has a dedicated data register in its internal memory for real-time caching of the physical extraction data of the high-square flat screen products. To quantify the actual grinding and crushing effect of the first-stage grinding mill and the overall system performance, the arithmetic logic unit within the closed-loop feedback unit calls floating-point division instructions to algebraically couple the acquired fine powder mass flow rate with the coarse powder mass flow rate.

[0089] The closed-loop feedback unit performs the following continuous mathematical calculation formula for the actual mass flow rate ratio: ; in, This is the current sampling time; The current sampling time The corresponding actual mass flow rate ratio; The current sampling time The corresponding fine powder mass flow rate; The current sampling time The corresponding coarse powder mass flow rate; A positive real constant is used to prevent program overflow caused by a division denominator of zero.

[0090] The actual mass flow rate ratio generated by the computational control module using the aforementioned continuous calculation formula reflects, at a physical level, the comprehensive scraping and extraction rate of the material after undergoing mechanical extrusion and aerodynamic flow field intervention. The closed-loop feedback unit stores the continuously calculated actual mass flow rate ratio in the process data image area of ​​the programmable logic controller. This actual mass flow rate ratio will serve as the core feedback basis for calculating the steady-state deviation of the main loop of the system in subsequent steps.

[0091] S302, the closed-loop feedback unit in the operation and control module performs proportional-integral-differential operations in the main loop based on the system's preset target mass flow ratio and actual mass flow ratio to generate a comprehensive feedback control command.

[0092] The closed-loop feedback unit stores target mass flow rates pre-set by process engineers for different flour varieties in its internal storage area. Each time feedback data is received, the arithmetic logic unit within the closed-loop feedback unit performs an algebraic difference between the target mass flow rate and the actual mass flow rate at the current sampling time, calculating the flow rate deviation that characterizes the current production process as deviating from the ideal process state.

[0093] Due to the slow physical accumulation of wear and tear on the grinding rollers and the mechanical lag of multiple pieces of equipment in the flour production process, the feedforward bias control of the pre-processing unit alone cannot completely eliminate the steady-state error generated during long-term operation. To correct this steady-state error, the closed-loop feedback unit calls the proportional-integral-derivative (PID) control algorithm library inside the programmable logic controller (PLC) and uses the calculated flow ratio deviation as the sole excitation source for mathematical calculation of the closed-loop control.

[0094] The closed-loop feedback unit specifically invokes discrete position proportional-integral-derivative (PID) operations for the comprehensive feedback control command. In the aforementioned closed-loop solution logic, the proportional part is responsible for physically amplifying and correcting the current flow ratio deviation of the system proportionally; the integral part mathematically accumulates the flow ratio deviations from all historical sampling times to forcibly eliminate the long-standing static offset of the system; and the derivative part calculates the rate of change of the deviation between two adjacent sampling times to suppress dynamic overshoot generated by the system in pursuit of the target mass-flow ratio. The comprehensive feedback control command generated through the above mathematical operations quantifies, at the physical level, the total amount of comprehensive action compensation that the control system must output at the current sampling time to restore the high-square screen output product to the ideal grinding effect.

[0095] S303, the closed-loop feedback unit in the operation and control module constructs the critical judgment condition for the anti-condensation temperature difference, and combines the underlying thermodynamic state data to execute the conditional branch logic of the generated comprehensive feedback control command.

[0096] The deviation in the flow rate ratio from the high-efficiency flat screen feedback objectively requires the control system to output a compensating action to restore the target grinding extraction rate. However, if the system blindly assigns the calculated integrated feedback control command directly to the mechanical execution unit to enhance the physical scraping intensity of the grinding rollers, the additional mechanical work will inevitably induce secondary vaporization of internal moisture when the gas phase in the workshop environment approaches the water vapor saturation critical point. This physical action conflict will cause the pipeline to cross the condensation boundary, resulting in the powdered grinding product absorbing moisture, agglomerating, and clogging the pneumatic conveying pipeline. To avoid this thermodynamic physical conflict, the closed-loop feedback unit must call upon the data cached in the previous state reconstruction unit to quantitatively assess the safety boundary of the current gas phase environment.

[0097] The arithmetic logic unit inside the closed-loop feedback unit retrieves variables from the global data block, executes the critical judgment condition for anti-condensation temperature difference, and calculates the algebraic difference between the ambient temperature and the dew point temperature at the current sampling time, i.e., the thermodynamic temperature difference margin at the current sampling time.

[0098] The closed-loop feedback unit compares the calculated thermodynamic temperature difference margin with the system's preset safe temperature difference threshold stored in memory using a digital comparator. When the thermodynamic temperature difference margin at the current sampling moment is less than or equal to the system's preset safe temperature difference threshold, it means that the actual ambient temperature is approaching the condensation point of water vapor. The closed-loop feedback unit sets a high humidity danger flag in the status register of the programmable logic controller, thereby logically determining that the internal flow field of the production equipment is in a high humidity critical condition.

[0099] Conversely, when the thermodynamic temperature difference margin corresponding to the current sampling moment is strictly greater than the system's preset safe temperature difference threshold, it means that the flow field has sufficient physical space to accommodate water vapor. The closed-loop feedback unit resets the high humidity danger flag in the status register, determining that the flow field inside the production equipment is currently in a dry and safe operating condition. This conditional branch algorithm achieves early interception of physical action conflict risks at the digital control level and specifies a unique logical path for the routing of subsequent control commands.

[0100] S304. The closed-loop feedback unit in the operation and control module, based on the state of the high humidity danger flag, superimposes the comprehensive feedback control command onto the pre-bias command using different mathematical weighting rules, and generates the final terminal action command sent to each execution unit.

[0101] The closed-loop feedback unit reads the high humidity danger flag in the status register and executes two independent control command routing and allocation algorithms.

[0102] When the high humidity hazard flag is in the reset state, it is determined that the current internal flow field of the equipment is in a dry and safe operating condition, and the physical environment allows for increased mechanical work without the risk of large-scale water vapor condensation. The closed-loop feedback unit directs the comprehensive feedback control command corresponding to the current sampling moment to the mechanical execution unit in the collaborative execution module, and performs linear superposition calculation with the pre-bias command of the mechanical physical domain, directly correcting the flow ratio deviation by enhancing the mechanical physical scraping force.

[0103] The mapping formula for the terminal action commands specifically executed by the closed-loop feedback unit under dry and safe operating conditions is as follows: ; ; ; ; in, This is the current sampling time; The current sampling time The corresponding final grinding roller spacing command; The preset basic value for the grinding roller spacing of the system; The current sampling time The corresponding grinding roller spacing offset; Assign weight coefficients to the spacing feedback; The current sampling time The corresponding integrated feedback control command; The current sampling time The corresponding final grinding roller speed difference command; The preset base value for the roller speed difference in the system; The current sampling time The corresponding roller speed difference offset; Assign weighting coefficients to the speed difference feedback; The current sampling time The corresponding final command given to the variable frequency pneumatic control valve; The preset aerodynamic opening value for the system; The current sampling time The corresponding variable frequency pneumatic control valve pre-opening command; The current sampling time The corresponding final proportional flow control valve command; The preset thermal control opening value for the system; The current sampling time The corresponding proportional flow control valve inverse proportional opening command.

[0104] When the high humidity danger flag is in the set state, the internal flow field of the equipment is determined to be in a critical high humidity condition. The closed-loop feedback unit actively freezes the distribution of comprehensive feedback control commands to the mechanical execution unit to avoid additional mechanical extrusion work that could induce secondary evaporation of moisture. Under the condition of limiting mechanical work, the closed-loop feedback unit distributes the comprehensive feedback control commands across the physical domain to the pneumatic and thermal control execution units in the collaborative execution module. By drastically increasing the pneumatic dehumidification and suction airflow and the internal cooling water flow rate, the surface temperature of the raw wheat material is forcibly reduced and the flow field characteristics are changed. Utilizing the increased brittleness of the material under low-temperature drying flow field conditions, the steady-state error of grinding efficiency is indirectly compensated.

[0105] The mapping formula for the terminal action command specifically executed by the closed-loop feedback unit under high humidity critical conditions is as follows: ; ; ; ; in, This is the current sampling time; The current sampling time The corresponding final grinding roller spacing command; The preset basic value for the grinding roller spacing of the system; The current sampling time The corresponding grinding roller spacing offset; The current sampling time The corresponding final grinding roller speed difference command; The preset base value for the roller speed difference in the system; The current sampling time The corresponding roller speed difference offset; The current sampling time The corresponding final command given to the variable frequency pneumatic control valve; The preset aerodynamic opening value for the system; The current sampling time The corresponding variable frequency pneumatic control valve pre-opening command; Assign weighting coefficients to the aerodynamic feedback; The current sampling time The corresponding integrated feedback control command; The current sampling time The corresponding final proportional flow control valve command; The preset thermal control opening value for the system; The current sampling time The corresponding proportional flow control valve inverse proportional opening command; Assign weighting coefficients to thermal control feedback.

[0106] The computation control module converts the final action command generated by the above distribution and calculation into a standard voltage or current signal to drive the actuator through the internally configured digital and analog output channels, and sends it to the collaborative execution module in parallel to realize decoupling and safe anti-collision control in multiple physical domain states.

[0107] Before sending various instructions to the collaborative execution module, the closed-loop feedback unit in the S305 operation and control module performs physical boundary limit constraint calculations on all control instructions to prevent the actuator from exceeding the limit and causing equipment damage, and simultaneously completes the closed-loop update of the control cycle.

[0108] Because the final control commands calculated theoretically may exceed the physical limits of the underlying mechanical equipment when encountering extreme physical conditions or sudden changes in feedback signals. For example, if the roller spacing command is too small, it may cause direct physical collision between the fast and slow rollers; if the opening command of the pneumatic valve or water valve exceeds its full range, it may cause the motor of the actuator to stall and overload. To ensure the long-term mechanical safety of the control system, the closed-loop feedback unit performs physical boundary limiting constraint calculations on the underlying actuators. This means that the terminal output command corresponding to the current sampling moment is uniformly limited to the physical lower limit threshold and physical upper limit threshold of the corresponding execution channel preset by the system. The terminal output command specifically refers to the final roller spacing command, the final roller speed difference command, the final variable frequency pneumatic regulating valve command, or the final proportional flow regulating valve command generated by the previous calculation stage. After completing the above physical boundary limiting constraints, the closed-loop feedback unit writes the limited terminal safety output commands of all channels into the physical output image register of the programmable logic controller. The collaborative execution module reads the above-mentioned safety-constrained instructions and drives the position servo motor, frequency converter, frequency converter pneumatic regulating valve and proportional flow regulating valve to perform physical displacement and frequency changes, thereby completing the multi-dimensional physical space reshaping of the mechanical crushing and extrusion pressure, shearing force, gas phase forced convection dehumidification and solid phase heat conduction cooling characteristics inside the first grinding mill.

[0109] As commands are issued and executed, the internal hardware clock of the computational control module triggers the start of the next control cycle, and continuously collects the latest mass flow rate of the high-square flat screen discharge pipeline from the fluid sensing unit, forming a continuous dynamic adjustment loop with interdependent beginnings and ends. This loop ensures that the overall scraping and extraction rate of the raw wheat material always converges stably within the target range set by the process, and reduces the risk of condensation and blockage inside the pneumatic conveying pipeline by relying on state condition branches.

[0110] The underlying detection and conversion of industrial-grade microwave powder flow sensors, the conventional tuning rules of PID control parameters, the implementation of PLC structured text or ladder diagram programming in accordance with international standards, the digital-to-analog conversion electrical control of industrial-grade analog output modules, and the underlying scheduling mechanism of PLC general cyclic scanning program architecture involved in the above S30 steps are all well-known technologies in this field. Those skilled in the art can implement them using conventional hardware or technical means, and will not be elaborated here.

[0111] Specific application examples: To better understand the technical solution of this invention, the invention will be further described in detail below with reference to specific application scenarios and accompanying drawings. This specific application embodiment is constructed in the first-stage milling section of a large flour processing plant processing Jimai 22 hard winter wheat. The system operates within a control system that includes a multi-dimensional sensing module and a collaborative execution module, and is responsible for acquiring basic physical data and directionally allocating action commands.

[0112] During system operation, the electromagnetic sensing unit in the multi-dimensional sensing module acquires the original impedance sequence according to a preset scanning time sequence. The parameters are set as follows: The current sampling time is set. Take a specific value and obtain the current sampling time. The corresponding low-frequency AC impedance sequence Ω is 300Ω, obtain the current sampling time. Corresponding high-frequency AC impedance sequence The current impedance is 150Ω. The state reconstruction unit in the arithmetic control module receives the above data and substitutes it into the formula to calculate the current sampling time. Corresponding impedance ratio sequence : ; The numerical calculation process is as follows: ; The calculation results show that at the current sampling time Corresponding impedance ratio sequence The value is 0.5. The environmental sensing unit in the multi-dimensional sensing module obtains the current sampling time. Corresponding ambient temperature The temperature is 25℃. The parameters are set as follows: The constant coefficients in the first constant coefficient matrix are pre-calibrated for a specific wheat variety. 10, constant coefficient 0.2, constant coefficient =1; the constant coefficients in the second constant coefficient matrix are the constant coefficients pre-calibrated for specific wheat varieties. 6. Constant coefficient 0.1, constant coefficient The value is 1.5. The state reconstruction unit uses the above parameters and substitutes them into the polynomial reconstruction equation to calculate the current sampling time. Corresponding cortical toughness characteristic value With the current sampling time Corresponding endosperm compressive strength characteristic value : ; ; The numerical calculation process is as follows: ; ; The calculation results show that the current sampling time obtained by the state reconstruction unit is Corresponding cortical toughness characteristic value The value is 11.0, and the current sampling time is... Corresponding endosperm compressive strength characteristic value It is 7.0.

[0113] The pre-processing unit in the arithmetic control module is equipped with a time compensation queue, which allows the system to acquire material conveying delay time. The corresponding span is used to obtain the material reaching the grinding chamber at the delayed sampling time. Corresponding cortical toughness characteristic value The value is 11.0, and the sampling time is delayed. Corresponding endosperm compressive strength characteristic value Version 7.0. The environmental sensing unit acquires the current sampling time. Corresponding grinding roller surface temperature The temperature is 35℃. The first weighting coefficients of the pre-defined multivariate prediction equation. The second weighting coefficient of the multivariate prediction equation is 1.2. The third weighting coefficient of the multivariate prediction equation is 0.8. The value is 0.5. The pre-processing unit substitutes the values ​​into the multivariate prediction equation to calculate the current sampling time. Corresponding phase change vaporization prediction index : ; The numerical calculation process is as follows: ; The calculation results show that at the current sampling time Corresponding phase change vaporization prediction index The value is 23.8. The preprocessing unit performs an algebraic mapping based on this exponent to generate the current sampling time. Corresponding roller spacing offset The value is 0.05, and the current sampling time. Corresponding variable frequency pneumatic control valve pre-opening command It is 10.0.

[0114] The closed-loop feedback unit in the arithmetic control module performs discrete position proportional-integral-differential operations to generate the current sampling time. Corresponding integrated feedback control command The value is 4.0. The closed-loop feedback unit calculates the thermodynamic temperature difference margin corresponding to the current sampling time as 1.5℃, and simultaneously reads the system's preset safe temperature difference critical threshold as 3.0℃. Since the thermodynamic temperature difference margin of 1.5℃ corresponding to the current sampling time is less than or equal to the system's preset safe temperature difference critical threshold of 3.0℃, the closed-loop feedback unit sets the high humidity danger flag in the status register, determining that the internal flow field of the equipment is currently in a high humidity critical condition. In this state, the closed-loop feedback unit actively freezes the allocation of integrated feedback control commands to the mechanical execution unit, distributing them across the physical domain to the pneumatic execution unit in the collaborative execution module. The parameter settings are as follows: Set the system's preset basic value for the grinding roller spacing. The value is 0.5, which is the system's preset basic value for aerodynamic opening. The aerodynamic feedback allocation weighting coefficient is 30. The value is 5.0. The closed-loop feedback unit substitutes the values ​​into the formula to calculate the current sampling time. The corresponding final grinding roller spacing command With the current sampling time The corresponding final variable frequency pneumatic control valve command : ; ; The numerical calculation process is as follows: ; ; The calculation results show that at the current sampling time The corresponding final grinding roller spacing command Maintained at 0.55, at the current sampling time. The corresponding final variable frequency pneumatic control valve command The value is 60.0. The control system limits the enhancement range of the mechanical extrusion action and directs the compensation work to the pneumatic dehumidification action.

[0115] To verify the stability of the control command orientation allocation logic in handling high humidity critical conditions, the system compared and output the operating data of multiple control algorithms under the same experimental environment. The total experiment duration was 120 minutes, with steam introduced into the system at the 60th minute to simulate changes in ambient humidity.

[0116] See attached document Figure 3 The horizontal axis represents the running time, and the vertical axis represents the thermodynamic temperature difference margin. In the first 60 minutes of the test, the temperature difference value extracted by the traditional control algorithm fluctuated around 5.2℃. After the 60th minute, the thermodynamic temperature difference margin curve representing the current sampling time extracted by the system of this invention shows a trend of crossing the system's preset safe temperature difference threshold of 3.0℃ and establishing a steady state around 1.5℃. This result verifies that the sensing module has the ability to extract flow field state data and provide boundary judgment conditions.

[0117] Based on the verification of boundary judgment conditions, the system compared the instruction output actions of different underlying devices. (See attached document) Figure 4 The horizontal axis represents the running time, and the vertical axis represents the final grinding roller spacing command on the left and the final variable frequency pneumatic control valve command on the right, respectively. During the first 60 minutes of the test, the final grinding roller spacing command fluctuated. After 60 minutes, the final grinding roller spacing command of this invention remained at a horizontal straight line of 0.55. The curve representing the final variable frequency pneumatic control valve command of this invention shows a trend of rising to around 60 and then fluctuating. This result verifies that the computational control module has the ability to avoid physical action conflicts.

[0118] To demonstrate the control reliability of this invention in actual processing, the system statistically analyzed the actual mass flow rate ratio under different strategies. (See attached diagram.) Figure 5 The horizontal axis represents running time, and the vertical axis represents the actual mass flow rate ratio. The actual mass flow rate ratio of the traditional control system oscillates after 60 minutes and tends to decrease to 0.85. The actual mass flow rate ratio of the control system of this invention shows a trend of recovery and maintenance at the target mass flow rate ratio of 1.2 after a brief transition. This result verifies that the present invention has significant technical advantages in avoiding equipment condensation and clogging while maintaining the actual mass flow rate ratio.

Claims

1. An intelligent control system for flour production process parameters, characterized in that, include: The multi-dimensional sensing module is used to acquire basic physical data in the production section and define the physical monitoring boundaries of the intelligent control system for flour production process parameters. The operation and control module is used to receive the physical basic data acquired by the multi-dimensional sensing module as raw data, convert the raw data into feature values ​​characterizing the moisture gradient of the material and the dew point temperature, estimate the amount of vaporization generated by mechanical work based on the feature values ​​characterizing the moisture gradient of the material, generate an instruction bias based on the amount of vaporization, calculate the flow ratio deviation of the fluid mass based on the raw data, perform proportional-integral-differential operation of the main loop to generate a deviation instruction, generate thermodynamic boundary conditions based on the dew point temperature, and directionally allocate the deviation instruction based on the thermodynamic boundary conditions. A collaborative execution module is used to receive the instruction bias and the deviation instruction, and to perform physical actions according to the instruction bias and the deviation instruction.

2. The intelligent control system for flour production process parameters according to claim 1, characterized in that, The multi-dimensional perception module and the collaborative execution module specifically include: The multi-dimensional sensing module includes an electromagnetic sensing unit arranged at the discharge end of the wheat silo, a fluid sensing unit arranged at the two outlets of the high-square flat screen, and an environmental sensing unit arranged in the grinding chamber of the first leather mill and in the pipeline of the production workshop. The physical data includes the low-frequency AC impedance sequence and the high-frequency AC impedance sequence obtained by the electromagnetic sensing unit, the coarse powder mass flow rate and the fine powder mass flow rate obtained by the fluid sensing unit, and the grinding roller surface temperature, ambient temperature and relative humidity obtained by the environmental sensing unit. The collaborative execution module includes a mechanical execution unit, a pneumatic execution unit, and a thermal control execution unit; The mechanical actuator is used to adjust the difference between the grinding roller spacing and the grinding roller speed of the first grinding mill. The pneumatic actuator includes a frequency conversion pneumatic regulating valve for adjusting the exhaust volume and air pressure of the grinding chamber suction pipe; The thermal control actuator includes a proportional flow regulating valve for adjusting the amount of cooling water entering the grinding roller.

3. The intelligent control system for flour production process parameters according to claim 2, characterized in that, The operation control module includes: The state reconstruction unit is used to convert the low-frequency AC impedance sequence, the high-frequency AC impedance sequence, the ambient temperature and the relative humidity into cortical toughness characteristic value, endosperm compressive strength characteristic value and the dew point temperature. The cortical toughness characteristic value and the endosperm compressive strength characteristic value constitute the characteristic value characterizing the moisture gradient of the material. A pre-processing unit is used to estimate the amount of vaporization generated by mechanical work based on the cortical toughness characteristic value and the endosperm compressive strength characteristic value, and to generate the instruction bias amount based on the amount of vaporization. The closed-loop feedback unit is used to calculate the flow ratio deviation based on the mass flow rate of the coarse powder and the mass flow rate of the fine powder, perform proportional-integral-differential operations in the main loop to generate the deviation command, generate the thermodynamic boundary conditions based on the ambient temperature and the dew point temperature, and allocate the deviation command in a directional manner based on the thermodynamic boundary conditions.

4. The intelligent control system for flour production process parameters according to claim 3, characterized in that, The state reconstruction unit is specifically used for: The impedance ratio sequence is constructed by quotienting the low-frequency AC impedance sequence and the high-frequency AC impedance sequence. Using the impedance ratio sequence and the ambient temperature as input variables, the cortical toughness characteristic value and the endosperm compressive strength characteristic value are calculated through a polynomial reconstruction equation. The ambient temperature and relative humidity are mathematically coupled to calculate the dew point temperature.

5. The intelligent control system for flour production process parameters according to claim 3, characterized in that, The pre-processing unit is specifically used for estimating the amount of vaporization generated by mechanical work in the following ways: Calculate the temperature difference between the surface temperature of the grinding roller and the ambient temperature; The cortex toughness characteristic value, the endosperm compressive strength characteristic value and the temperature difference are linearly weighted and coupled to generate a phase change vaporization prediction index. The phase change vaporization prediction index is used as the vaporization amount.

6. The intelligent control system for flour production process parameters according to claim 5, characterized in that, The pre-processing unit is specifically used to generate the instruction bias based on the vaporization amount as follows: Using the phase change vaporization prediction index, the roller spacing offset corresponding to the extrusion force and the roller speed difference offset corresponding to the shear force are calculated respectively through algebraic scaling operations; Using the phase change vaporization prediction index, an algebraic mapping equation with quadratic amplification characteristics is used to generate a pre-opening command for the frequency conversion pneumatic control valve. Using the phase change vaporization prediction index, an inverse proportional opening command for the proportional flow control valve is generated through a mapping equation based on an inverse proportional algebraic structure. The instruction offset includes the roller spacing offset, the roller speed difference offset, the variable frequency pneumatic regulating valve pre-opening instruction, and the proportional flow regulating valve inverse proportional opening instruction.

7. The intelligent control system for flour production process parameters according to claim 3, characterized in that, The closed-loop feedback unit is specifically used for: calculating the flow ratio deviation and generating the deviation command in the following ways: The actual mass flow rate ratio is calculated by algebraically coupling the mass flow rate of the fine powder with the mass flow rate of the coarse powder. The flow ratio deviation is calculated by performing an algebraic difference between the preset target mass flow ratio and the actual mass flow ratio. The flow ratio deviation is used as the excitation source for closed-loop control to perform discrete position proportional-integral-derivative operations to generate a comprehensive feedback control command. The integrated feedback control command is used as the deviation command.

8. The intelligent control system for flour production process parameters according to claim 3, characterized in that, The closed-loop feedback unit is specifically used to generate the thermodynamic boundary conditions as follows: Calculate the algebraic difference between the ambient temperature and the dew point temperature to obtain the thermodynamic temperature margin; The thermodynamic temperature difference margin is compared with the preset safe temperature difference critical threshold. When the thermodynamic temperature difference margin is less than or equal to the safety temperature difference critical threshold, the high humidity danger flag is set in the status register; When the thermodynamic temperature difference margin is greater than the safe temperature difference critical threshold, the high humidity danger flag bit is reset in the status register. The state of the high humidity hazard marker constitutes the thermodynamic boundary condition.

9. The intelligent control system for flour production process parameters according to claim 8, characterized in that, The closed-loop feedback unit is specifically used for: When allocating the deviation command in a directional manner according to the thermodynamic boundary conditions, it is used for: When the high humidity hazard flag is in the reset state, the closed-loop feedback unit directs the deviation command to the mechanical execution unit and calculates the deviation command by superimposing it with the command bias amount corresponding to the mechanical execution unit. When the high humidity hazard flag is in the set state, the closed-loop feedback unit freezes and assigns the deviation command to the mechanical actuator, and assigns the deviation command to the pneumatic actuator and the thermal control actuator, and calculates the deviation command by superimposing it with the command offset amount corresponding to the pneumatic actuator and the thermal control actuator.

10. The intelligent control system for flour production process parameters according to claim 9, characterized in that, The closed-loop feedback unit is also equipped with a safety timer; When the high humidity hazard flag is in the set state, the closed-loop feedback unit starts the safety timer to keep track of the time. When the cumulative duration of the safety timer exceeds the preset limit safety threshold, the closed-loop feedback unit generates a system shutdown command and sends the system shutdown command to the collaborative execution module. The collaborative execution module stops executing the physical action according to the system shutdown command.

Images