Kitchen food processing production control method and system based on internet of things technology

Abstract

This invention relates to the field of control and regulation system technology, specifically to a method and system for controlling food processing production based on Internet of Things (IoT) technology. The method connects a thermal imaging sensor and a microwave moisture sensor via an IoT edge gateway to acquire images of the heat distribution in the heating area and sequences of material moisture content. These sequences are then input into a thermodynamic and moisture evaporation coupling model to calculate the dynamic specific heat capacity and latent heat of phase change, and combined with the thermal conductivity coefficient to predict the thermal inertia time constant. Based on the time constant, a feedforward adjustment is calculated and superimposed on a proportional-integral-derivative feedback adjustment generated based on temperature deviation to form a comprehensive control signal. This signal controls the heating actuator to adjust its output power in advance before the temperature reaches a set threshold. This invention eliminates the thermal inertia and temperature overshoot caused by dynamic changes in the material's specific heat capacity, reduces the temperature fluctuation amplitude within the heating chamber, and avoids defects such as localized charring or insufficient heating of the material.

Description

Technical Field

[0001] This invention discloses a method and system for controlling food processing and production based on Internet of Things (IoT) technology, which relates to the field of control and regulation system technology. Background Technology

[0002] Existing food processing control systems typically employ a single contact temperature sensor within the heating chamber. The controller acquires the material temperature measured by this sensor, calculates the feedback adjustment based on the deviation between the set temperature and the real-time temperature using a proportional-integral-derivative (PID) control algorithm, and adjusts the output power of the heating equipment accordingly. However, in actual processing, different batches of food materials exhibit variations in initial temperature and moisture content, and moisture evaporation continues throughout the heating process.

[0003] Because contact temperature sensors can only detect the surface temperature of a localized area of ​​the material, changes in this measurement lag significantly behind the actual heat exchange state caused by the latent heat absorbed by moisture evaporation within the material. When the moisture content of the material changes, its specific heat capacity changes accordingly. The system cannot predict the change in the material's thermal inertia in advance and can only initiate power correction after the material's surface temperature has exceeded the set threshold. This control mechanism based on a single hysteresis feedback produces temperature overshoot and control lag, causing temperature fluctuations within the heating chamber to exceed the allowable range, leading to defects such as localized charring or uneven heating of the material. Summary of the Invention

[0004] The purpose of this invention is to provide a method and system for controlling the production of food processing in the kitchen based on Internet of Things (IoT) technology, so as to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows: A method for controlling food processing and production based on Internet of Things (IoT) technology includes: connecting a thermal imaging sensor and a microwave moisture sensor through an IoT edge gateway to collect thermal distribution image sequences and material moisture content sequences of the heating area in the kitchen. The heat distribution image sequence and the material moisture content sequence are input into the thermodynamic and moisture evaporation coupling model built into the IoT edge gateway to calculate the current dynamic specific heat capacity and latent heat of phase change of the material. By combining the inherent thermal conductivity of the heating equipment with the dynamic specific heat capacity and the latent heat of phase change, the thermal inertia time constant under the current operating conditions is predicted through the thermodynamic and moisture evaporation coupling model. The feedforward adjustment is calculated based on the predicted thermal inertia time constant, and the feedforward adjustment is superimposed on the proportional-integral-derivative feedback adjustment generated based on the deviation between the target temperature and the real-time temperature to form a comprehensive control signal. The integrated control signal is sent to the heating actuator, which controls the heating actuator to adjust its output power in advance according to the integrated control signal before the material temperature reaches the set threshold.

[0006] Preferably, the step of acquiring the heat distribution image sequence and the material moisture content sequence of the kitchen heating area includes: obtaining the image acquisition frame rate of the thermal imaging sensor and the sampling frequency of the microwave moisture sensor; using the timestamp of the image acquisition frame rate as a reference, interpolating and resampling the moisture data acquired by the microwave moisture sensor to align the moisture data with the heat distribution image on the time axis. In the spatial dimension, a subset of pixels corresponding to the coverage area of ​​the microwave moisture sensor beam in the thermal distribution image is extracted, the average temperature of each pixel in the subset is calculated as the average temperature value, and the average temperature value is bound to the moisture data at the corresponding time after interpolation and resampling to generate a multimodal synchronous data frame sequence containing a unified timestamp, spatial location coordinates, the average temperature value and the moisture data.

[0007] Preferably, the step of calculating the current dynamic specific heat capacity and latent heat of phase change of the material includes: for consecutive data frames in the multimodal synchronous data frame sequence, calculating the rate of moisture decrease and the average temperature gradient between adjacent data frames; The moisture descent rate and the average temperature gradient are input into a preset physical property parameter mapping matrix, wherein the physical property parameter mapping matrix is ​​constructed based on the material's internal moisture migration equation and energy conservation equation. By solving the partial differential relationship in the physical property parameter mapping matrix, the sensible heat absorption component used to compensate for the temperature rise and the latent heat absorption component used for the water state transformation are separated. The dynamic specific heat capacity is determined based on the sensible heat absorption component, and the latent heat of phase change is determined based on the latent heat absorption component.

[0008] Preferably, the step of predicting the thermal inertia time constant under the current operating condition includes: obtaining the heat source radiation attenuation coefficient of the heating device under the current output power and the thermal convection boundary conditions inside the heating cavity; The heat source radiation attenuation coefficient and the heat convection boundary condition are introduced as constraint variables into the first-order inertial transfer function, and the dynamic specific heat capacity and the latent heat of phase change are used as dynamic parameters of the first-order inertial transfer function. By solving the time response expression of the first-order inertial transfer function, which includes the constraint variables and the dynamic parameters, the time scale corresponding to when the temperature change curve reaches a preset ratio is extracted, and the time scale is determined as the thermal inertial time constant.

[0009] Preferably, the step of calculating the feedforward adjustment amount based on the predicted thermal inertia time constant includes: using the difference between the real-time temperature at the current moment and the set threshold as the target temperature difference for feedforward control, and constructing a negative exponential decay function with respect to the thermal inertia time constant; The negative exponential decay function is integrated within the target temperature difference range to obtain the power reduction area required to eliminate the temperature overshoot caused by the thermal inertia time constant. Based on the ratio of the current actual output power of the heating actuator to the power reduction area, the power reduction area is converted into a power compensation sequence that decreases over time, and the power compensation sequence is used as the feedforward adjustment amount.

[0010] Preferably, the step of superimposing the feedforward adjustment amount into the proportional-integral-derivative feedback adjustment amount includes: calculating the integral cumulative value in the proportional-integral-derivative feedback adjustment amount; Determine the sign state of the integral cumulative value and the feedforward adjustment amount. When the integral cumulative value and the feedforward adjustment amount are both negative and the absolute value of the feedforward adjustment amount is greater than the preset intervention threshold, perform reverse attenuation processing on the integral cumulative value according to the slope of the power compensation sequence. The integral cumulative value after reverse attenuation processing is recombined with the proportional and derivative terms to generate the proportional-integral-derivative feedback adjustment amount that suppresses integral saturation. Then, the proportional-integral-derivative feedback adjustment amount is linearly superimposed with the feedforward adjustment amount.

[0011] Preferably, the step of extracting a subset of pixels in the thermal distribution image corresponding to the coverage area of ​​the microwave moisture sensor beam further includes: performing edge gradient detection on the thermal distribution image to identify the physical boundary contour of the heating tray; Based on the physical boundary contour, the edge pixel regions in the thermal distribution image that exceed the physical boundary contour of the tray are cropped out, wherein the edge pixel regions contain ambient radiation interference heat. When calculating the average temperature gradient, the gradient is solved only within the effective region inside the cropped heat distribution image to eliminate the interference of environmental radiation heat on the calculation process of the sensible heat absorption component.

[0012] Preferably, the step of obtaining the thermal convection boundary conditions inside the heating cavity includes: arranging an array of temperature sensors on the inner wall surface of the heating cavity to collect the real-time wall surface temperature distribution; Based on the real-time wall temperature distribution and the material surface temperature in the multimodal synchronous data frame sequence, the dynamic temperature difference field between the inner wall and the material surface is calculated. Substituting the dynamic temperature difference field into the Nusselt number correlation, the local convective heat transfer coefficient inside the heating cavity that varies with spatial position is obtained. The mean value of the local convective heat transfer coefficient is then used as the thermal convection boundary condition and input into the first-order inertial transfer function.

[0013] Preferably, before the step of determining the sign state of the integral accumulation value and the feedforward adjustment amount, the method further includes: reading the material type identifier of the current batch of kitchen materials, and querying the standard thermal conductivity and standard moisture evaporation rate corresponding to the material type identifier in a preset material thermal property database; Calculate the inherent thermal response delay ratio corresponding to the material type identifier based on the standard thermal conductivity and the standard moisture evaporation rate; The preset intervention threshold is scaled and adjusted using the inherent thermal response delay ratio so that the preset intervention threshold for different material types matches the inherent thermal conductivity characteristics of the material.

[0014] The food processing and production control system based on Internet of Things (IoT) technology includes a thermal imaging sensor, a microwave moisture sensor, an IoT edge gateway, and a heating actuator. The thermal imaging sensor and the microwave moisture sensor are both communicatively connected to the IoT edge gateway to collect thermal distribution image sequences and material moisture content sequences of the heating area of ​​the kitchen, respectively. The IoT edge gateway has a built-in thermodynamic and moisture evaporation coupling model. The IoT edge gateway is used to input the heat distribution image sequence and the material moisture content sequence into the thermodynamic and moisture evaporation coupling model to calculate the current dynamic specific heat capacity and latent heat of phase change of the material. Combined with the inherent thermal conductivity coefficient of the heating equipment, the thermal inertia time constant under the current operating condition is predicted. The feedforward adjustment amount is calculated based on the thermal inertia time constant and superimposed on the proportional-integral-derivative feedback adjustment amount to form a comprehensive control signal. The heating actuator is communicatively connected to the IoT edge gateway to receive the integrated control signal and adjust the output power in advance according to the integrated control signal before the material temperature reaches the set threshold.

[0015] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. This invention acquires multimodal data by connecting a thermal imaging sensor and a microwave moisture sensor through an IoT edge gateway. It uses a thermodynamic and moisture evaporation coupling model to calculate dynamic specific heat capacity and latent heat of phase change to predict the thermal inertia time constant. Based on this, a feedforward adjustment is generated and superimposed on the proportional-integral-derivative feedback adjustment, enabling the heating actuator to adjust its output power in advance before the material temperature reaches a set threshold. This technique reconstructs hysteresis feedback control into a hybrid feedforward and feedback control based on changes in material properties. This eliminates thermal inertia and temperature overshoot caused by dynamic changes in the material's specific heat capacity, reduces temperature fluctuations within the heating chamber, and avoids coking defects caused by localized overheating or undercooking defects caused by insufficient heating.

[0016] 2. The solution improves the accuracy of dynamic specific heat capacity and latent heat of phase change calculations by interpolating and resampling moisture data based on the image acquisition frame rate and combining it with spatial pixel subset extraction, thus aligning multimodal data in the spatiotemporal dimensions. By introducing a local convective heat transfer coefficient when solving the first-order inertial transfer function, the predicted thermal inertia time constant is made to better reflect the actual cavity environment. Integral saturation of the controller is suppressed by reverse attenuation of the integral cumulative value in the proportional-integral-derivative feedback control quantity based on the power compensation sequence. Interference from environmental radiation heat on the calculation of sensible heat absorption components is eliminated by cropping edge pixel regions that exceed the physical boundary contour of the tray. The intervention threshold of the feedforward control quantity is adjusted by scaling the inherent thermal response delay corresponding to the material type identifier, ensuring that the feedforward control logic matches the inherent thermal conductivity characteristics of different materials. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the overall process of the food processing and production control method based on Internet of Things technology of the present invention; Figure 2 This is a flowchart illustrating the multimodal data spatiotemporal alignment method of the present invention; Figure 3 This is a flowchart illustrating the method for calculating dynamic specific heat capacity and latent heat of phase change according to the present invention. Figure 4 This is a flowchart illustrating the thermal inertia time constant prediction method of the present invention; Figure 5 This is a schematic flowchart of the feedforward adjustment calculation and integral saturation suppression method of the present invention; Figure 6 This is a flowchart illustrating the edge pixel cropping and dynamic threshold adjustment method of the present invention. Detailed Implementation

[0018] This invention relates to the field of food thermal processing control technology, specifically to a food processing production control method and system based on Internet of Things (IoT) technology.

[0019] Please refer to Figure 1 This embodiment provides a method for controlling food processing production based on Internet of Things (IoT) technology, implemented using an IoT edge gateway. The IoT edge gateway establishes bidirectional data interaction channels with a thermal imaging sensor, a microwave moisture sensor, and a heating actuator via wired or wireless communication links. The communication links employ a low-latency, high-reliability transmission protocol to ensure that both the uplink transmission of sensor data and the downlink transmission of control commands meet real-time requirements. The thermal imaging sensor's field of view completely covers the spatial range of the heating area, ensuring that the entire material-bearing area within the heating region is within its field of view. The beam coverage area of ​​the microwave moisture sensor matches the material-bearing area in the thermal imaging sensor's field of view, ensuring that the moisture data collected by the microwave moisture sensor corresponds to the same material area as the temperature data collected by the thermal imaging sensor. After the heating equipment starts operating, the IoT edge gateway sends synchronous acquisition commands to the thermal imaging sensor and the microwave moisture sensor. The thermal imaging sensor continuously acquires thermal distribution images of the heating area according to a preset image acquisition frame rate. Each pixel in each frame of the thermal distribution image corresponds to the temperature value at the corresponding spatial location within the heating area. The thermal imaging sensor arranges the acquired continuous thermal distribution images in chronological order, generating a thermal distribution image sequence and uploading it to the IoT edge gateway. A microwave moisture sensor continuously collects moisture content data of materials within a heating area at a preset sampling frequency. Each sample corresponds to the average moisture content within the material's beam coverage area at the sampling time. The microwave moisture sensor arranges the collected continuous moisture content data in chronological order, generating a material moisture content sequence, and uploads it to an IoT edge gateway. The IoT edge gateway locally caches the received thermal distribution image sequence and material moisture content sequence. During caching, each image frame and each moisture sampling data point is marked with a corresponding acquisition timestamp to ensure the chronological order of the sequence data is traceable.

[0020] The IoT edge gateway inputs cached heat distribution image sequences and material moisture content sequences into a built-in thermodynamic and moisture evaporation coupling model. This model is constructed based on the law of energy conservation and the internal moisture migration patterns of the material. The solution process is completed locally on the IoT edge gateway, without relying on cloud computing resources, thus avoiding control delays caused by data transmission. The core control equation of the thermodynamic and moisture evaporation coupling model is based on the energy balance relationship of the material during heating, and its specific expression is:

[0021] In the formula, For the density of the material, For the volume of the material, This refers to the dynamic specific heat capacity of a material as a function of temperature. This refers to the real-time temperature of the material. For time variables, The latent heat of phase change of a material as temperature changes. This represents the real-time moisture content of the material. The heat flow rate input from the heating equipment to the material. This represents the heat loss from the material to the heating chamber environment. A coupled thermodynamic and moisture evaporation model analyzes the heat distribution image sequence frame by frame, extracting the temperature values ​​of the material region in each frame to generate a temperature change sequence of the material over continuous time. Simultaneously, it analyzes the material moisture content sequence to generate a moisture change sequence over continuous time. The model substitutes the temperature and moisture change sequences into the core control equations and discretizes them using the inherent thermal conductivity coefficient of the heating equipment. The discretization process employs a fixed time step, consistent with the image acquisition cycle of the thermal imaging sensor, ensuring synchronization between the solution and data acquisition processes. During discretization, the model separates the total heat absorbed by the material into two parts: a sensible heat absorption component for raising the material temperature and a latent heat absorption component for the phase change of moisture evaporation within the material. By solving these two components separately, the dynamic specific heat capacity and latent heat of phase change of the material at the current moment are calculated. The dynamic specific heat capacity directly corresponds to the sensible heat absorption component, and the latent heat of phase change directly corresponds to the latent heat absorption component.

[0022] After acquiring the dynamic specific heat capacity and latent heat of phase change of the material at the current moment, the IoT edge gateway, combined with the inherent thermal conductivity of the heating equipment, predicts the thermal inertia time constant under the current operating conditions through a thermodynamic and moisture evaporation coupling model. The thermal inertia time constant is a core parameter characterizing the temperature response lag of the heating system; its magnitude directly determines the lag time of temperature change relative to heating power adjustment. The thermodynamic and moisture evaporation coupling model incorporates a calculation model for the thermal inertia time constant, and the expression of the calculation model is as follows:

[0023] In the formula, The thermal inertia time constant, This refers to the equivalent specific heat capacity of the material. The inherent thermal conductivity coefficient of the heating equipment, The heat exchange area between the material and the heat source is defined as follows: The equivalent specific heat capacity of the material is the sum of the equivalent specific heat capacity corresponding to the dynamic specific heat capacity and the latent heat of phase change. The equivalent specific heat capacity corresponding to the latent heat of phase change is calculated using the latent heat of phase change and the rate of change of moisture content during the material's temperature change process. Converting the latent heat of phase change to the equivalent specific heat capacity unifies the dimensions of thermophysical parameters and simplifies the calculation process of the thermal inertia time constant. The inherent thermal conductivity coefficient of the heating equipment is a fixed parameter calibrated before the equipment leaves the factory. It characterizes the inherent ability of the heating equipment's heat source to transfer heat to the material. Its value is stored in the local storage unit of the IoT edge gateway and can be retrieved according to the operating conditions of the heating equipment. The model substitutes the calculated dynamic specific heat capacity, latent heat of phase change, and the retrieved inherent thermal conductivity coefficient into the above calculation model to obtain the thermal inertia time constant under the current operating conditions.

[0024] The IoT edge gateway calculates the feedforward adjustment amount based on the predicted thermal inertia time constant. This feedforward adjustment amount compensates for the thermal inertia hysteresis of the heating system, ensuring that the power adjustment of the heating actuator occurs before the material temperature reaches the set threshold, thus preventing overshoot. The calculation of the feedforward adjustment amount is based on the thermal inertia time constant and the current temperature state of the material. The difference between the current real-time temperature of the material and the set threshold is used as the calculation basis. The set threshold is the upper limit of the target temperature that the material needs to reach during the heating process. When the material temperature reaches the set threshold, the heating actuator needs to stop or reduce its output power to prevent the temperature from rising further. In the calculation of the feedforward adjustment amount, a decay function related to the thermal inertia time constant is constructed. The total power to be reduced in advance is obtained through the integral operation of the decay function. This total power is then converted into a continuous adjustment amount that varies with time, serving as the final feedforward adjustment amount.

[0025] While calculating the feedforward adjustment, the IoT edge gateway acquires the target temperature and real-time temperature of the material within the heating area. The target temperature is the final heating temperature required by the heating process for the current batch of material, and the real-time temperature is calculated from the average temperature of the material area in the current frame of the heat distribution image sequence. The IoT edge gateway calculates the deviation between the target temperature and the real-time temperature, and generates a proportional-integral-derivative (PID) feedback adjustment based on the deviation. The expression for the PID feedback adjustment is as follows:

[0026] In the formula, for Proportional-integral-derivative feedback adjustment at any given time. This is the proportional gain coefficient. This is the integral gain coefficient. The differential gain coefficient, for The deviation between the target temperature and the real-time temperature. The proportional gain, integral gain, and derivative gain coefficients are pre-tuned control parameters stored in the local storage unit of the IoT edge gateway and can be adjusted according to the type of material and the heating process. The proportional term is used to quickly generate an adjustment amount based on the current deviation value, the integral term is used to eliminate the steady-state error of the system, and the derivative term is used to generate an adjustment amount in advance based on the rate of change of the deviation value, suppressing rapid changes in the deviation value.

[0027] The IoT edge gateway superimposes the calculated feedforward adjustment into the proportional-integral-derivative feedback adjustment to form a comprehensive control signal. The superposition process uses a linear superposition method, and the specific expression is as follows:

[0028] In the formula, for The integrated control signal at any given time, for The feedforward adjustment amount at any given time. During the linear superposition process, the amplitude of the integrated control signal is limited. The upper and lower limits of the limit correspond to the rated output power range of the heating actuator to prevent the generated integrated control signal from exceeding the adjustable range of the heating actuator and to ensure the executability of the control commands.

[0029] The IoT edge gateway sends the generated integrated control signal to the heating actuator via a pre-established communication link. Upon receiving the integrated control signal, the heating actuator analyzes it and converts it into a corresponding power drive signal, controlling the internal power conversion circuit to adjust the output power. The heating actuator's power adjustment is completed before the material temperature reaches the set threshold. The adjustment magnitude and timing are jointly determined by the feedforward and feedback adjustment values ​​in the integrated control signal. The feedforward adjustment value determines the lead time and total magnitude of the power adjustment, while the feedback adjustment value dynamically corrects the power adjustment magnitude based on real-time temperature deviations.

[0030] In this embodiment, the core parameters of the thermodynamic and water evaporation coupled model are defined as shown in the table below: Table 1. Definition of core parameters for the coupled thermodynamic and water evaporation model. ; Table 1 clarifies the physical meaning, data source, and value constraints of all core parameters in the coupled thermodynamic and water evaporation model, ensuring the reproducibility of the model solution process and the consistency of the calculation results. This allows those skilled in the art to define and completely reproduce the model solution process based on these parameters.

[0031] This embodiment achieves local acquisition and processing of multi-source sensor data through an IoT edge gateway. Based on a coupled thermodynamic and moisture evaporation model, it realizes real-time calculation of dynamic thermophysical parameters of materials and accurate prediction of thermal inertia time constant. By linearly superimposing feedforward and feedback adjustment quantities to form a comprehensive control signal, it realizes advance adjustment of the output power of the heating actuator, eliminates the temperature overshoot problem caused by thermal inertia lag of the heating system, avoids local coking of materials due to temperature overshoot, and avoids uneven heating of materials due to insufficient heating.

[0032] In a preferred embodiment, reference Figure 2 When collecting heat distribution image sequences and material moisture content sequences in the kitchen heating area, the IoT edge gateway first obtains the image acquisition frame rate of the thermal imaging sensor and the sampling frequency of the microwave moisture sensor. The image acquisition frame rate corresponds to the number of heat distribution image frames acquired per second by the thermal imaging sensor, and the sampling frequency corresponds to the number of moisture content data points acquired per second by the microwave moisture sensor. The IoT edge gateway uses the timestamp corresponding to the image acquisition frame rate of the thermal imaging sensor as a benchmark to interpolate and resample the moisture data acquired by the microwave moisture sensor, aligning the moisture data with the heat distribution images on the time axis. Since the image acquisition frame rate of the thermal imaging sensor and the sampling frequency of the microwave moisture sensor typically differ (the sampling frequency of the microwave moisture sensor is generally higher than the image acquisition frame rate of the thermal imaging sensor), the number of moisture sampling data points within the same time period exceeds the number of frames in the heat distribution images. The interpolation and resampling process matches the timestamp of each heat distribution image frame with the corresponding moisture content data, ensuring that each heat distribution image frame has moisture content data corresponding to the same moment. The interpolation and resampling uses a linear interpolation algorithm, where the timestamp corresponding to each image frame in the heat distribution image sequence is used for interpolation. The IoT edge gateway searches for data from microwave moisture sensors that are related to... Two adjacent sampling times and ,in Obtain moisture content data at two sampling times. and Calculated through linear interpolation Moisture content data at time 1 The expression for linear interpolation is:

[0033] In the formula, For timestamps The corresponding interpolated and resampled moisture content data, Less than And with The closest sampling time of the microwave moisture sensor greater than And with The closest sampling time of the microwave moisture sensor for Moisture content sampling value at time 10:00 for Moisture content samples at each time point. Through linear interpolation resampling, all timestamps corresponding to the thermal distribution images have matching moisture content data, achieving precise alignment between thermal distribution data and moisture data in the time dimension.

[0034] After aligning the timeline, the IoT edge gateway processes the thermal distribution image spatially, extracting a subset of pixels corresponding to the coverage area of ​​the microwave moisture sensor beam. The IoT edge gateway first performs edge gradient detection on the thermal distribution image to identify the physical boundary contour of the heating tray. This process involves calculating the gradient magnitude of each pixel in the thermal distribution image, obtained by the temperature difference between the pixel and its neighboring pixels. The temperature gradient magnitude at the physical boundary of the heating tray is significantly higher than the temperature gradient magnitude between the material area inside the tray and the external environment. The image is segmented using a preset gradient threshold to identify boundary pixels. A contour fitting algorithm is then used to fit these boundary pixels, obtaining the closed physical boundary contour of the heating tray. Based on the fitted physical boundary contour, the IoT edge gateway cropps the edge pixel regions in the thermal distribution image that extend beyond the tray's physical boundary contour. These edge pixel regions correspond to non-material areas such as the inner wall of the heating cavity and the support structure, and contain ambient radiant heat. The cropped thermal distribution image retains only the effective material area inside the tray, eliminating the interference of ambient radiant heat on the material temperature calculation.

[0035] After cropping the thermal distribution image, the IoT edge gateway extracts a subset of pixels corresponding to the beam coverage area of ​​the microwave moisture sensor within the cropped effective area. The beam coverage area of ​​the microwave moisture sensor is a fixed circular or rectangular region, and its corresponding position and size in the thermal distribution image are determined through pre-completion spatial calibration. The spatial calibration process maps the beam coverage area of ​​the microwave moisture sensor to the pixel coordinate system of the thermal distribution image, obtaining the corresponding pixel coordinate range. Based on the mapped pixel coordinate range, the IoT edge gateway extracts all pixels in the corresponding region of the thermal distribution image, forming a pixel subset. It then calculates the average temperature of all pixels within this subset as the average temperature value of the region. The IoT edge gateway binds the calculated average temperature value with the moisture data at the corresponding time after interpolation and resampling, generating a multimodal synchronization data frame. All multimodal synchronization data frames are arranged in chronological order to form a multimodal synchronization data frame sequence.

[0036] In this embodiment, the field structure of the multimodal synchronization data frame is shown in the following table: Table 2 Field Structure Table of Multimodal Synchronization Data Frames ; Table 2 clarifies the complete field structure of the multimodal synchronization data frame, ensuring the accurate binding of thermal distribution data and moisture data in the spatiotemporal dimension. This provides input data in a unified format for the subsequent calculation of dynamic thermal property parameters, enabling those skilled in the art to generate and parse the multimodal synchronization data frame based on this field structure.

[0037] Based on the generated multimodal synchronous data frame sequence, the IoT edge gateway calculates the current dynamic specific heat capacity and latent heat of phase change of the material. For consecutive data frames in the multimodal synchronous data frame sequence, the IoT edge gateway calculates the moisture decay rate and average temperature gradient between adjacent data frames. Adjacent data frames are two data frames with consecutive timestamps in the sequence, namely the [number]th [frame], ... Frame and the The frames and their corresponding timestamps are as follows: and The corresponding moisture contents are respectively and The corresponding average temperature values ​​are respectively and The formula for calculating the rate of water loss is:

[0038] In the formula, The rate of moisture loss between adjacent data frames. For the first The moisture content corresponding to the frame data. For the first The moisture content corresponding to the frame data. For the first Timestamps of frame data For the first The timestamp of the frame data. The expression for calculating the average temperature gradient is:

[0039] In the formula, The average temperature gradient between adjacent data frames. For the first The average temperature value corresponding to the frame data. For the first The average temperature value corresponding to the frame data.

[0040] The IoT edge gateway inputs the calculated moisture decay rate and average temperature gradient into a preset physical property parameter mapping matrix. This matrix, a two-dimensional transformation matrix, is constructed based on the material's internal moisture migration equation and energy conservation equation. The elements of the matrix are obtained by fitting experimental data of the material's thermophysical properties. The inputs are the moisture decay rate and average temperature gradient, and the outputs are sensible heat absorption components and latent heat absorption components. The IoT edge gateway separates the sensible heat absorption component (for compensating for temperature rise) and the latent heat absorption component (for moisture state transformation) by solving the partial differential relationships in the physical property parameter mapping matrix. The sensible heat absorption component represents the unit heat required for the material's temperature to rise and directly corresponds to the material's dynamic specific heat capacity. The latent heat absorption component represents the unit heat required for liquid water to transform into gaseous water and directly corresponds to the material's latent heat of phase change. The IoT edge gateway determines the material's current dynamic specific heat capacity based on the sensible heat absorption component and its current latent heat of phase change based on the latent heat absorption component. When calculating the average temperature gradient, the IoT edge gateway only performs gradient calculations within the effective region of the cropped heat distribution image, eliminating the interference of environmental radiation heat on the calculation process of the sensible heat absorption component.

[0041] This embodiment achieves spatiotemporal synchronization and alignment of thermal distribution data and moisture data through interpolation resampling in the time dimension and pixel subset extraction in the spatial dimension. It eliminates environmental radiation interference through edge gradient detection and image cropping, and achieves accurate separation of sensible heat component and latent heat component through physical property parameter mapping matrix. This improves the calculation accuracy of dynamic specific heat capacity and phase change latent heat, and provides a reliable parameter basis for the prediction of thermal inertia time constant.

[0042] In another preferred embodiment, reference Figure 3 When predicting the thermal inertia time constant under current operating conditions, the IoT edge gateway first obtains the heat source radiation attenuation coefficient of the heating device at its current output power and the thermal convection boundary conditions inside the heating cavity. The heat source radiation attenuation coefficient characterizes the degree of attenuation of heat radiated from the heat source during its propagation within the heating cavity. Its value is related to the reflectivity of the inner wall of the heating cavity, the cavity space dimensions, and the placement of materials. Different output powers of the heating device correspond to different heat source radiation intensities, and thus different heat source radiation attenuation coefficients. The mapping relationship between the heat source radiation attenuation coefficient and the output power of the heating device is obtained through pre-completed cavity calibration experiments. This mapping relationship is stored in the local storage unit of the IoT edge gateway. Based on the current output power of the heating device, the IoT edge gateway calls the corresponding mapping relationship to obtain the heat source radiation attenuation coefficient at the current output power.

[0043] The process by which the IoT edge gateway acquires the thermal convection boundary conditions inside the heating cavity is as follows: An array of temperature sensors is arranged on the inner wall of the heating cavity. Each sensing unit of the array is uniformly distributed along the inner wall, covering all surfaces. Each sensing unit collects the inner wall temperature at its corresponding installation location, generating real-time wall temperature distribution data and uploading it to the IoT edge gateway. The IoT edge gateway acquires the real-time wall temperature distribution data collected by the array of temperature sensors, and simultaneously acquires the material surface temperature data from the multimodal synchronous data frame sequence. The material surface temperature data is calculated from the average temperature of the surface pixels in the material region of the thermal distribution image. The IoT edge gateway calculates the temperature difference between the positions of each sensing unit on the inner wall and the corresponding positions on the material surface, generating a dynamic temperature difference field between the inner wall and the material surface. This dynamic temperature difference field characterizes the temperature distribution differences within the heating cavity and is the core factor driving the convection heat transfer of air inside the cavity.

[0044] The IoT edge gateway substitutes the generated dynamic temperature difference field into the Nusselt number correlation to obtain the local convective heat transfer coefficient inside the heating cavity that varies with spatial location. The Nusselt number correlation is a dimensionless correlation characterizing the intensity of convective heat transfer, and its expression is:

[0045] In the formula, For Nusselt numbers, , , The experimental constants related to the structure of the heating cavity were obtained through cavity convection heat transfer experiments. Re is the Reynolds number, which characterizes the flow state of the fluid inside the cavity. The Prandtl number characterizes the physical properties of the fluid inside the cavity. The Reynolds number is calculated using the air velocity, kinematic viscosity, and characteristic length of the cavity, while the Prandtl number is obtained from a table of air temperature and physical property parameters inside the cavity. After obtaining the Nusselt number, the IoT edge gateway calculates the local convective heat transfer coefficient using the formula for the convective heat transfer coefficient:

[0046] In the formula, The local convective heat transfer coefficient is... The thermal conductivity of the air inside the cavity is obtained by referring to the physical property table using the average temperature of the air inside the cavity. Let be the characteristic length of the heating cavity, and be a fixed structural parameter of the heating cavity. The IoT edge gateway calculates the mean of the local convective heat transfer coefficient corresponding to all spatial locations inside the heating cavity, and inputs this mean as the thermal convection boundary condition into the first-order inertial transfer function.

[0047] The IoT edge gateway incorporates the acquired heat source radiation attenuation coefficient and thermal convection boundary conditions as constraint variables into the first-order inertial transfer function, and uses the calculated dynamic specific heat capacity and latent heat of phase change as dynamic parameters of the first-order inertial transfer function. The first-order inertial transfer function characterizes the temperature response characteristics of the heating system; its input is the output heat flow of the heating equipment, and its output is the temperature change of the material. The expression of the transfer function is in Laplace domain form:

[0048] In the formula, Let be the first-order inertial transfer function of the heating system. For the Laplace operator, The static gain of the system is determined by the radiation attenuation coefficient of the heat source and the thermal convection boundary conditions. The thermal inertia time constant is determined by both the dynamic specific heat capacity and the latent heat of phase change. The system's static gain... It characterizes the proportional relationship between the steady-state temperature response of the heating system and the input heat flow. Its value is negatively correlated with the radiation attenuation coefficient of the heat source and negatively correlated with the convective heat transfer coefficient corresponding to the thermal convection boundary conditions.

[0049] The IoT edge gateway extracts the time scale corresponding to when the temperature change curve reaches a preset proportion by solving the time response expression of the first-order inertial transfer function, which includes constraint variables and dynamic parameters. This time scale is then defined as the thermal inertial time constant. The unit step response expression of the first-order inertial transfer function is as follows: This expression characterizes the change in material temperature over time under a step-input heat flow in the heating system. When the temperature change reaches 63.2% of the steady-state value, the corresponding time... The numerical value of thermal inertia time constant Since the values ​​are equal, the IoT edge gateway extracts the corresponding time scale by solving the time response expression, which is used as the thermal inertia time constant under the current operating conditions.

[0050] In this embodiment, the calculation parameters for the local convective heat transfer coefficient of the heating cavity are shown in the table below: Table 3. Calculation parameters for local convective heat transfer coefficient of heating cavity ; Table 3 clarifies the calculation methods and data sources for all parameters in the calculation process of the convective heat transfer coefficient, ensuring the calculation accuracy of the thermal convection boundary conditions and making the predicted results of the thermal inertia time constant consistent with the actual operating environment of the heating cavity. This allows those skilled in the art to complete the complete calculation process of the convective heat transfer coefficient based on this parameter table.

[0051] This embodiment uses an array of temperature sensors to collect the temperature distribution on the inner wall of the heating cavity, solves the local convective heat transfer coefficient based on the Nusselt number correlation, obtains the thermal convection boundary conditions that fit the actual working conditions, and introduces the heat source radiation attenuation coefficient and the thermal convection boundary conditions as constraints into the first-order inertial transfer function to achieve accurate prediction of the thermal inertial time constant, providing accurate hysteresis characteristic parameters for the calculation of feedforward regulation.

[0052] In yet another preferred embodiment, reference is made to Figures 4 to 6 When the IoT edge gateway calculates the feedforward adjustment based on the predicted thermal inertia time constant, it uses the difference between the current real-time temperature and a set threshold as the target temperature difference for feedforward control. The set threshold is the upper limit of temperature that cannot be exceeded during material heating, and the target temperature difference is the difference between the set threshold and the current real-time temperature, representing the temperature rise required for the material temperature to reach the set threshold. The IoT edge gateway constructs a negative exponential decay function with respect to the thermal inertia time constant. This function characterizes the temperature response characteristics of the heating system under thermal inertia. The decay rate of the function is determined by the thermal inertia time constant; the larger the thermal inertia time constant, the slower the decay rate of the function, and the more pronounced the system's hysteresis characteristics. The IoT edge gateway integrates the negative exponential decay function over the time interval corresponding to the target temperature difference to obtain the power reduction area required to eliminate temperature overshoot caused by the thermal inertia time constant. The power reduction area represents the total time integral of the heating power that needs to be reduced in advance to avoid temperature overshoot. The IoT edge gateway converts the power reduction area into a time-decreasing power compensation sequence based on the ratio of the current actual output power of the heating actuator to the power reduction area. Each value in the power compensation sequence corresponds to the power adjustment at a given moment. The time length of the sequence is consistent with the thermal inertia time constant. The IoT edge gateway uses this power compensation sequence as the feedforward adjustment quantity. The expression for calculating the feedforward adjustment quantity is:

[0053] In the formula, for Feedforward adjustment at time intervals This represents the current actual output power of the heating actuator. The power-temperature conversion coefficient corresponds to the target temperature difference, representing the change in material temperature per unit change in power. For the target temperature difference, The predicted thermal inertia time constant, This is the integral time variable. The negative sign in the expression indicates that the feedforward adjustment is a power reduction measure, used to reduce the output power of the heating actuator in advance to prevent the material temperature from exceeding the set threshold.

[0054] When adding the feedforward regulation to the proportional-integral-derivative (PID) feedback regulation, the IoT edge gateway first calculates the integral cumulative value in the PFD. This integral cumulative value is the integral of the deviation between the target temperature and the real-time temperature over time, used to eliminate steady-state errors in the system. As the material temperature approaches a set threshold, the deviation gradually decreases. If the integral cumulative value is too large, it can lead to integral saturation, preventing the heating actuator's output power from timely adjusting and causing temperature overshoot. The IoT edge gateway determines the sign of the integral cumulative value and the feedforward regulation. When both are negative and the absolute value of the feedforward regulation is greater than a preset intervention threshold, the integral cumulative value is attenuated in reverse according to the slope of the power compensation sequence. The reverse attenuation process involves generating a corresponding reverse attenuation coefficient based on the decreasing slope of the power compensation sequence, multiplying the integral cumulative value by the reverse attenuation coefficient, and gradually decreasing the integral cumulative value over time to avoid integral saturation caused by an excessively large integral cumulative value. The IoT edge gateway recombines the integral cumulative value after reverse attenuation processing with the proportional and derivative terms to generate a proportional-integral-derivative feedback adjustment quantity that suppresses integral saturation. Then, it linearly superimposes this proportional-integral-derivative feedback adjustment quantity with the feedforward adjustment quantity to generate a comprehensive control signal.

[0055] Before determining the sign state of the integral accumulation value and the feedforward adjustment, the IoT edge gateway reads the material type identifier of the current batch of kitchen materials. The material type identifier is a unique code representing the material category; different types of kitchen materials correspond to different material type identifiers. The IoT edge gateway queries a pre-set material thermal property database for the standard thermal conductivity and standard moisture evaporation rate corresponding to the material type identifier. The material thermal property database stores standard thermal property parameters for different types of kitchen materials, including standard thermal conductivity, standard specific heat capacity, and standard moisture evaporation rate, which are obtained through experimental testing. Based on the retrieved standard thermal conductivity and standard moisture evaporation rate, the IoT edge gateway calculates the inherent thermal response delay ratio corresponding to the material type identifier. The inherent thermal response delay ratio characterizes the degree to which the thermal conductivity characteristics of different types of materials affect the system's thermal response. The lower the thermal conductivity and the faster the moisture evaporation rate of the material, the larger the inherent thermal response delay ratio, and the more obvious the thermal hysteresis characteristics of the system. The IoT edge gateway uses the inherent thermal response delay ratio to scale and adjust the preset intervention threshold. The scaling and adjustment method is to multiply the preset intervention threshold by the inherent thermal response delay ratio to obtain the adjusted intervention threshold. This makes the preset intervention threshold for different material types match the inherent thermal conductivity characteristics of the materials. For materials with more obvious thermal hysteresis characteristics, the intervention threshold is reduced, allowing the feedforward adjustment to intervene earlier and adjust the power in advance.

[0056] This embodiment also provides a food processing and production control system based on Internet of Things (IoT) technology. The system includes a thermal imaging sensor, a microwave moisture sensor, an IoT edge gateway, and a heating actuator. Both the thermal imaging sensor and the microwave moisture sensor are connected to the IoT edge gateway via a communication link. The thermal imaging sensor's field of view covers the heating area of ​​the food processing area, continuously acquiring thermal distribution images of the heating area, generating a thermal distribution image sequence, and uploading it to the IoT edge gateway. The microwave moisture sensor's beam coverage area matches the material-bearing area in the thermal imaging sensor's field of view, continuously acquiring moisture content data of the material within the heating area, generating a material moisture content sequence, and uploading it to the IoT edge gateway. The IoT edge gateway has a built-in thermodynamics and moisture evaporation coupling model. It receives the sequence data uploaded by the thermal imaging sensor and the microwave moisture sensor, inputs the thermal distribution image sequence and the material moisture content sequence into the thermodynamics and moisture evaporation coupling model, calculates the material's current dynamic specific heat capacity and latent heat of phase change, predicts the thermal inertia time constant under the current operating conditions based on the inherent thermal conductivity of the heating equipment, calculates the feedforward adjustment amount based on the thermal inertia time constant, and superimposes it into the proportional-integral-derivative feedback adjustment amount to form a comprehensive control signal. The heating actuator is connected to the IoT edge gateway via a communication link to receive the comprehensive control signal sent by the IoT edge gateway, convert the comprehensive control signal into a power drive signal, and adjust the output power in advance according to the comprehensive control signal before the material temperature reaches the set threshold.

[0057] In this embodiment, the intervention threshold scaling parameters corresponding to different material types are shown in the following table: Table 4. Intervention threshold scaling parameters for different material types ; Table 4 clarifies the thermophysical parameters, inherent thermal response delay ratio, and intervention threshold scaling factor corresponding to different types of materials. This ensures that the intervention timing of feedforward control matches the inherent thermal conductivity characteristics of different materials, improves the material adaptability of the control logic, and enables those skilled in the art to adjust the intervention threshold corresponding to different materials based on this parameter table.

[0058] This embodiment calculates the feedforward adjustment amount by integrating a negative exponential decay function based on the thermal inertia time constant, thereby achieving precise adjustment of heating power in advance. It suppresses the integral saturation phenomenon of the system by reverse decay processing of the integral accumulation value, and adjusts the feedforward intervention threshold by the thermophysical parameters corresponding to the material type identifier, so that the control logic can be adapted to different types of kitchen materials, further improving the stability and accuracy of heating control.

Claims

1. A method for controlling the processing and production of food in the kitchen based on Internet of Things (IoT) technology, characterized in that: include: By connecting thermal imaging sensors and microwave moisture sensors through an IoT edge gateway, the heat distribution image sequence and material moisture content sequence of the kitchen heating area are collected. The heat distribution image sequence and the material moisture content sequence are input into the thermodynamic and moisture evaporation coupling model built into the IoT edge gateway to calculate the current dynamic specific heat capacity and latent heat of phase change of the material. By combining the inherent thermal conductivity of the heating equipment with the dynamic specific heat capacity and the latent heat of phase change, the thermal inertia time constant under the current operating conditions is predicted through the thermodynamic and moisture evaporation coupling model. The feedforward adjustment is calculated based on the predicted thermal inertia time constant, and the feedforward adjustment is superimposed on the proportional-integral-derivative feedback adjustment generated based on the deviation between the target temperature and the real-time temperature to form a comprehensive control signal. The integrated control signal is sent to the heating actuator, which controls the heating actuator to adjust its output power in advance according to the integrated control signal before the material temperature reaches the set threshold.

2. The method for controlling the processing and production of food in the kitchen based on Internet of Things technology according to claim 1, characterized in that, The steps for collecting heat distribution image sequences and material moisture content sequences in the heating area of ​​a kitchen include: obtaining the image acquisition frame rate of the thermal imaging sensor and the sampling frequency of the microwave moisture sensor; using the timestamp of the image acquisition frame rate as a reference, interpolating and resampling the moisture data collected by the microwave moisture sensor to align the moisture data with the heat distribution image on the time axis. In the spatial dimension, a subset of pixels corresponding to the coverage area of ​​the microwave moisture sensor beam in the thermal distribution image is extracted, the average temperature of each pixel in the subset is calculated as the average temperature value, and the average temperature value is bound to the moisture data at the corresponding time after interpolation and resampling to generate a multimodal synchronous data frame sequence containing a unified timestamp, spatial location coordinates, the average temperature value and the moisture data.

3. The method for controlling the processing and production of food in the kitchen based on Internet of Things technology according to claim 2, characterized in that, The steps for calculating the current dynamic specific heat capacity and latent heat of phase change of the material include: for consecutive data frames in the multimodal synchronous data frame sequence, calculating the rate of moisture decrease and the average temperature gradient between adjacent data frames; The moisture descent rate and the average temperature gradient are input into a preset physical property parameter mapping matrix, wherein the physical property parameter mapping matrix is ​​constructed based on the material's internal moisture migration equation and energy conservation equation. By solving the partial differential relationship in the physical property parameter mapping matrix, the sensible heat absorption component used to compensate for the temperature rise and the latent heat absorption component used for the water state transformation are separated. The dynamic specific heat capacity is determined based on the sensible heat absorption component, and the latent heat of phase change is determined based on the latent heat absorption component.

4. The method for controlling the processing and production of food in the kitchen based on Internet of Things technology according to claim 3, characterized in that, The steps for predicting the thermal inertia time constant under the current operating conditions include: obtaining the heat source radiation attenuation coefficient of the heating device under the current output power and the thermal convection boundary conditions inside the heating cavity; The heat source radiation attenuation coefficient and the heat convection boundary condition are introduced as constraint variables into the first-order inertial transfer function, and the dynamic specific heat capacity and the latent heat of phase change are used as dynamic parameters of the first-order inertial transfer function. By solving the time response expression of the first-order inertial transfer function, which includes the constraint variables and the dynamic parameters, the time scale corresponding to when the temperature change curve reaches a preset ratio is extracted, and the time scale is determined as the thermal inertial time constant.

5. The method for controlling the processing and production of food in the kitchen based on Internet of Things technology according to claim 4, characterized in that, The step of calculating the feedforward adjustment amount based on the predicted thermal inertia time constant includes: using the difference between the real-time temperature at the current moment and the set threshold as the target temperature difference for feedforward control, and constructing a negative exponential decay function with respect to the thermal inertia time constant. The negative exponential decay function is integrated within the target temperature difference range to obtain the power reduction area required to eliminate the temperature overshoot caused by the thermal inertia time constant. Based on the ratio of the current actual output power of the heating actuator to the power reduction area, the power reduction area is converted into a power compensation sequence that decreases over time, and the power compensation sequence is used as the feedforward adjustment amount.

6. The method for controlling the processing and production of food in the kitchen based on Internet of Things technology according to claim 5, characterized in that, The step of superimposing the feedforward adjustment amount into the proportional-integral-derivative (PID) feedback adjustment amount includes: calculating the integral cumulative value in the PLD feedback adjustment amount; Determine the sign state of the integral cumulative value and the feedforward adjustment amount. When the integral cumulative value and the feedforward adjustment amount are both negative and the absolute value of the feedforward adjustment amount is greater than the preset intervention threshold, perform reverse attenuation processing on the integral cumulative value according to the slope of the power compensation sequence. The integral cumulative value after reverse attenuation processing is recombined with the proportional and derivative terms to generate the proportional-integral-derivative feedback adjustment amount that suppresses integral saturation. Then, the proportional-integral-derivative feedback adjustment amount is linearly superimposed with the feedforward adjustment amount.

7. The method for controlling the processing and production of food in the kitchen based on Internet of Things technology according to claim 3, characterized in that, The step of extracting a subset of pixels in the heat distribution image that corresponds to the coverage area of ​​the microwave moisture sensor beam further includes: performing edge gradient detection on the heat distribution image to identify the physical boundary contour of the heating tray; Based on the physical boundary contour, the edge pixel regions in the thermal distribution image that exceed the physical boundary contour of the tray are cropped out, wherein the edge pixel regions contain ambient radiation interference heat. When calculating the average temperature gradient, the gradient is solved only within the effective region inside the cropped heat distribution image to eliminate the interference of environmental radiation heat on the calculation process of the sensible heat absorption component.

8. The method for controlling the processing and production of food in the kitchen based on Internet of Things technology according to claim 4, characterized in that, The steps for obtaining the thermal convection boundary conditions inside the heating cavity include: arranging an array of temperature sensors on the inner wall of the heating cavity to collect the real-time wall temperature distribution; Based on the real-time wall temperature distribution and the material surface temperature in the multimodal synchronous data frame sequence, the dynamic temperature difference field between the inner wall and the material surface is calculated. Substituting the dynamic temperature difference field into the Nusselt number correlation, the local convective heat transfer coefficient inside the heating cavity that varies with spatial position is obtained. The mean value of the local convective heat transfer coefficient is then used as the thermal convection boundary condition and input into the first-order inertial transfer function.

9. The method for controlling the processing and production of food in the kitchen based on Internet of Things technology according to claim 6, characterized in that, Before determining the sign state of the integral cumulative value and the feedforward adjustment amount, the method further includes: reading the material type identifier of the current batch of kitchen materials, and querying the standard thermal conductivity and standard moisture evaporation rate corresponding to the material type identifier in a preset material thermal property database; Calculate the inherent thermal response delay ratio corresponding to the material type identifier based on the standard thermal conductivity and the standard moisture evaporation rate; The preset intervention threshold is scaled and adjusted using the inherent thermal response delay ratio so that the preset intervention threshold for different material types matches the inherent thermal conductivity characteristics of the material.

10. A food processing and production control system based on Internet of Things (IoT) technology, characterized in that: This includes thermal imaging sensors, microwave moisture sensors, IoT edge gateways, and heating actuators; Both the thermal imaging sensor and the microwave moisture sensor are communicatively connected to the IoT edge gateway, and are used to collect thermal distribution image sequences and material moisture content sequences of the kitchen heating area, respectively. The IoT edge gateway has a built-in thermodynamic and moisture evaporation coupling model. The IoT edge gateway is used to input the heat distribution image sequence and the material moisture content sequence into the thermodynamic and moisture evaporation coupling model to calculate the current dynamic specific heat capacity and latent heat of phase change of the material. Combined with the inherent thermal conductivity coefficient of the heating equipment, the thermal inertia time constant under the current operating condition is predicted. The feedforward adjustment amount is calculated based on the thermal inertia time constant and superimposed on the proportional-integral-derivative feedback adjustment amount to form a comprehensive control signal. The heating actuator is communicatively connected to the IoT edge gateway to receive the integrated control signal and adjust the output power in advance according to the integrated control signal before the material temperature reaches the set threshold.

Images