Real-time control method and system for hot air rotating furnace air speed based on sensor data

Abstract

The present application relates to the technical field of automatic control, in particular to a hot air rotary furnace air speed real-time control method and system based on sensor data. The method comprises: collecting the inlet and outlet air temperature and real-time humidity of the heating chamber and preprocessing; calculating the inlet and outlet air temperature difference and time integration, and mapping to obtain a process progress factor representing the baking physical stage through an S-type growth model; constructing a dynamic humidity constraint logic based on the humidity deviation and the process progress factor to obtain a humidity sensitivity coefficient; correcting the basic air speed based on the humidity sensitivity coefficient to generate a target air speed instruction. The present application builds a time-varying constraint model conforming to the physical change law of the dough, controls the air speed based on the humidity at the initial stage of baking to prevent skinning, and automatically releases the control at the later stage of baking to improve the heat exchange efficiency, effectively solving the problems of small product volume and uneven coloring caused by traditional constant air speed.

Description

Technical Field

[0001] This invention relates to the field of automatic control technology, and specifically to a method and system for real-time control of the wind speed of a hot air rotary furnace based on sensor data. Background Technology

[0002] Hot air rotary ovens are core equipment in the modern baking industry. Their working principle is to use a circulating fan to bring heat from the heating chamber into the oven cavity, and then exchange heat with the dough on the rotating rack through forced convection. During the baking process, the wind speed directly determines the efficiency of heat exchange and the rate of evaporation of moisture on the surface of the dough.

[0003] Existing hot air rotary oven control technology typically uses a fixed-speed motor or only provides a few speed settings for manual adjustment. This constant or simple step-by-step fan speed control method cannot adapt to the complex physicochemical changes of dough during baking. Specifically, it has the following main drawbacks: The contradiction between overprotection and insufficient heat: In the early stage of baking, that is, the dough expansion period, the dough needs to keep its surface moist to facilitate volume expansion. If the air speed is too high at this time, it will quickly remove the moisture from the surface of the dough, causing the surface to harden and form a crust prematurely, which will hinder the expansion of internal gases and result in a smaller finished product. In the later stage of baking, that is, the browning and caramelization period, the internal structure of the dough has been set and needs to be browned and crisped through the Maillard reaction. If a low air speed is used at this time in order to prevent hardening and crust formation, it will lead to low heat exchange efficiency, and heat will be unable to penetrate the center of the dough, resulting in uneven browning and an insufficiently crispy crust. Lack of real-time response to load changes: When the amount of dough baked in a single batch changes, the heat demand in the oven will also fluctuate drastically. The constant wind speed cannot detect this load change, resulting in undercooked dough when fully loaded and burnt dough when unloaded.

[0004] Therefore, there is a need for a real-time control method for the airflow of a hot air rotary oven that can sense the physical stages of baking and intelligently and dynamically switch between humidity constraints and efficiency pursuits. Summary of the Invention

[0005] To address the problem that existing hot air rotary oven control technologies cannot simultaneously balance humidity control and production efficiency at different stages of baking, this invention provides a method and system for real-time wind speed control of hot air rotary ovens based on sensor data.

[0006] In a first aspect, the present invention provides a method for real-time control of the air velocity in a hot air rotary furnace based on sensor data, employing the following technical solution: The temperature and humidity data of the heating chamber and furnace are collected in real time by a preset data acquisition unit and preprocessed to obtain real-time temperature and humidity sequences. The real-time temperature sequence includes the inlet and return air temperatures of the heating chamber, and the real-time humidity sequence includes the real-time humidity in the furnace. The difference between the inlet and return air temperatures is calculated to obtain the inlet and outlet air temperature difference. The inlet and outlet air temperature difference is integrated over time, and the integration result is mapped through a preset nonlinear mapping model to obtain a process progress factor. The difference between the real-time humidity and a preset benchmark value is calculated to obtain the humidity deviation. Based on the humidity deviation and the process progress factor, a humidity sensitivity coefficient is calculated. The humidity sensitivity coefficient is used to correct the preset base wind speed to generate a target wind speed command. The target wind speed command is used to drive the circulating fan to achieve real-time control of the wind speed of the hot air rotary furnace.

[0007] This invention constructs a process progress factor reflecting the physical maturity of dough by integrating the temperature difference between the inlet and outlet air over time. Then, it combines the humidity deviation to obtain a humidity sensitivity coefficient to correct the base wind speed. During the dough expansion period, it can suppress the wind speed to prevent surface crusting, and during the dough coloring and caramelization period, it can automatically release the constraint to improve heat exchange efficiency. While ensuring the maximum volume of the finished product, it achieves uniform surface color and crispness, effectively solving the problem that traditional constant wind speed control cannot simultaneously protect volume and improve efficiency.

[0008] Furthermore, the preset nonlinear mapping model is specifically an S-shaped growth model, and the time integration operation specifically includes: cumulatively calculating the change of the inlet and outlet air temperature difference over time to obtain the cumulative heat absorption of the dough; inputting the cumulative heat absorption of the dough as an independent variable into a preset logistic function model for nonlinear mapping, calculating a value between 0 and 1, which is defined as the process progress factor; wherein, the growth center of the logistic function model is determined by a preset heat threshold parameter, and the process progress factor increases with the increase of the cumulative heat absorption of the dough.

[0009] This invention uses a logistic function model based on heat accumulation to calculate the process progress factor. Compared with the traditional control logic based solely on time, this method uses the integral of the temperature difference between the inlet and outlet air to characterize the actual enthalpy change absorbed by the dough. It can automatically adapt to the baking requirements of dough at full load, half load, or different initial temperatures, ensuring that the control strategy is always synchronized with the actual degree of ripeness inside the dough, and avoiding false ripening or burning caused by load changes.

[0010] Further, the humidity sensitivity coefficient includes: constructing an attenuation weight term using the process progress factor to dynamically adjust the humidity deviation, thereby obtaining the humidity sensitivity coefficient; wherein, when the process progress factor approaches 0, the humidity sensitivity coefficient decreases as the humidity deviation increases; when the process progress factor approaches 1, the humidity sensitivity coefficient approaches 1.

[0011] This invention utilizes a process progress factor as a decay weight term to adjust the influence of humidity deviation, achieving a smooth soft switch from strict humidity protection in the early stage of baking to ignoring humidity constraints in the later stage of baking. This not only prevents turbulent impacts on the heat flow field in the heating chamber caused by sudden changes in control strategy, but also ensures from a mathematical logic perspective that the wind speed adjustment fully follows the physical law of moisture evaporation on the dough surface.

[0012] Furthermore, the target wind speed command includes: obtaining a preset base wind speed, constructing a positive gain compensation term using the inlet and outlet air temperature difference, multiplying the base wind speed by the positive gain compensation term to obtain an intermediate wind speed value, and multiplying the intermediate wind speed value by the humidity sensitivity coefficient to obtain the target wind speed command.

[0013] This invention introduces a positive gain compensation term based on the temperature difference between the inlet and outlet air during the generation of the target wind speed command. This allows for rapid compensation of heat loss, effectively suppressing drastic temperature fluctuations and ensuring the consistency of each batch of baked goods.

[0014] Furthermore, the preprocessing includes: denoising the collected temperature data to filter out noise, and simultaneously performing spatiotemporal alignment on the temperature data to obtain the real-time temperature sequence.

[0015] This invention eliminates the spatiotemporal misalignment of data caused by different sensor installation locations by introducing a time delay parameter, ensuring the correspondence between the inlet and return air temperatures used in the calculation, improving the physical authenticity and accuracy of the collected data, and providing a reliable data foundation for subsequent control algorithms.

[0016] Furthermore, the inlet air temperature is collected using a first temperature sensor located behind the outlet guide plate of the heating chamber, and the return air temperature is collected using a second temperature sensor located at the center of the return air inlet grille of the furnace.

[0017] Furthermore, the driving of the circulating fan specifically involves: converting the target wind speed command into a corresponding frequency control signal, sending the frequency control signal to the frequency converter controlling the circulating fan via a serial communication interface, and the frequency converter adjusting the output voltage frequency according to the frequency control signal, thereby changing the rotational speed of the circulating fan.

[0018] Furthermore, the calorie threshold parameter was determined through a comprehensive standard baking experiment.

[0019] Furthermore, the real-time wind speed control method for the hot air rotary furnace based on sensor data also includes: real-time monitoring of the rate of change of the real-time humidity, and when the rate of change exceeds a preset abnormal threshold, locking the target wind speed command to a preset safe idle speed value until the rate of change is lower than the abnormal threshold.

[0020] Secondly, this invention provides a real-time control system for the wind speed of a hot blast rotary furnace based on sensor data, employing the following technical solution: A real-time control system for the wind speed of a hot blast rotary furnace based on sensor data includes: a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the above-mentioned real-time control method for the wind speed of a hot blast rotary furnace based on sensor data is implemented. By adopting the above technical solution, the above-mentioned real-time wind speed control method for hot air rotary furnace based on sensor data is generated into a computer program and stored in a container, so that it can be loaded and executed by the processor. In this way, a terminal device can be made based on the memory and the processor for convenient use.

[0021] The present invention has the following technical effects: This invention comprehensively utilizes thermodynamic principles and automatic control technology. By constructing a process progress factor based on energy accumulation, it effectively solves the problem of wind speed contradiction faced by hot air rotary ovens in complex baking processes. That is, low wind speed is required in the early stage of baking to retain moisture and promote expansion, while high wind speed is required in the later stage to promote heat exchange and promote browning. This invention uses the time integral of the temperature difference between the inlet and outlet air to evaluate the maturity of the dough, so that it is no longer limited to a fixed time program. It can adaptively adjust when faced with interference such as voltage fluctuations and changes in ambient temperature. Furthermore, through a nonlinear dynamic weight adjustment mechanism, it achieves smooth adjustment of the control strategy between different process stages, avoiding overshoot oscillations caused by step adjustment. This improves energy utilization efficiency while ensuring the color and texture of the product. This invention integrates multiple advanced algorithms such as signal spatiotemporal alignment, thermal load feedforward compensation, and S-shaped nonlinear mapping. It can not only effectively suppress system errors caused by sensor installation position, but also respond quickly when the temperature fluctuates drastically. This highly intelligent control system reduces the dependence on the professional skills of operators and provides strong technical support for the standardization and large-scale production of the baking industry. Attached Figure Description

[0022] Figure 1 This is a flowchart of a real-time wind speed control method for a hot air rotary furnace based on sensor data, provided in an embodiment of the present invention. Figure 2This is an adaptive control curve diagram of the entire baking process provided in the embodiments of the present invention. Detailed Implementation

[0023] This invention provides a method for real-time control of the air speed in a hot blast rotary furnace based on sensor data, referring to... Figure 1 This includes steps S1-S4: S1: Data Acquisition and Preprocessing.

[0024] Specifically, this step aims to acquire raw data of the internal thermodynamic environment of the hot air rotary furnace through a high-frequency, high-precision sensor network, and eliminate noise interference and spatiotemporal asynchrony through digital signal processing technology, so as to provide reliable physical input for subsequent control algorithms.

[0025] 1. Hardware environment construction and sensor deployment This embodiment uses an industrial-grade embedded microcontroller based on the ARM Cortex-M4 core and equipped with a floating-point arithmetic unit as the central processing unit. The core calculation and logic control tasks of the control system in this embodiment are undertaken by the central processing unit. A data acquisition unit is set up, including a temperature acquisition unit and a humidity acquisition unit. The temperature acquisition unit includes a first temperature sensor and a second temperature sensor. In this embodiment, a PT100 platinum resistance thermometer with Class A accuracy is used in conjunction with a high-speed probe with a thermal response time of less than 3 seconds. The humidity acquisition unit is a humidity sensor. In this embodiment, a capacitive humidity transmitter that can withstand high temperatures of 300℃ is used. Considering the special physical structure of the hot air rotary furnace, the layout of the acquisition unit is as follows: (1) First temperature sensor: installed 10cm behind the baffle at the outlet of the heating chamber. The airflow has been mixed but has not yet come into contact with the dough. The baffle effectively blocks the direct infrared radiation from the outside, ensuring that the temperature of the hot airflow is collected rather than the radiation temperature. (2) Second temperature sensor: installed at the geometric center of the furnace return air vent grille, where the return gas after flowing through all the dough is gathered, and its temperature can objectively reflect the average state after heat exchange; (3) Humidity sensor: installed in the static pressure zone on the side wall of the furnace to avoid reading fluctuations caused by direct hot air blowing.

[0026] 2. Data Acquisition and Filtering The data acquisition unit collects temperature and humidity data in real time and uploads them to the central processing unit. The sampling frequency is set to 20Hz, which corresponds to the sampling period. The collected data is processed by moving average filtering. In this embodiment, the length of the moving window is 10. After filtering, the temperature collected by the first temperature sensor is defined as the inlet air temperature, denoted as... The temperature collected by the second temperature sensor is defined as the return air temperature, denoted as... The humidity collected by the humidity sensor is defined as the real-time humidity, denoted as ; For time The real-time humidity is recorded as... The real-time humidity data is sorted by time and grouped into a set, which is named the real-time humidity sequence.

[0027] 3. Spatiotemporal alignment of data Because there is a 2-3 meter distance between the first and second temperature sensors, it takes a certain amount of time for hot air to travel from the inlet to the outlet, i.e., a transmission delay. If the difference between the inlet and outlet temperatures collected at the same moment is directly used for calculation, it is essentially subtracting the residual old heat generated in the past from the newly generated heat. During baking, especially in transient processes where the oven temperature fluctuates drastically due to frequent start-stop of the burner or opening of the oven door, this spatiotemporal misalignment will cause serious distortion in the calculated temperature difference, thus misleading the control algorithm. Therefore, it is necessary to perform spatiotemporal alignment on the collected temperature data. The specific steps are as follows: (1) During the equipment commissioning stage, the hot air rotary furnace is placed in the cold furnace state, the heating system is turned off, and the circulating fan is controlled to run continuously at the rated speed until the airflow circulation in the hot air rotary furnace reaches the fluid steady state. (2) When the heating system is turned on at full power by manual control, the air inlet temperature will rise rapidly. After a period of time, the hot air will reach the return air inlet, causing the return air temperature to rise. (3) The curves of the inlet air temperature and the return air temperature change with time are recorded synchronously using a high-frequency data logger. The moment when the curve of the inlet air temperature begins to change abruptly and the moment when the curve of the return air temperature begins to change abruptly are extracted respectively. The difference between the two moments when the curves begin to change abruptly is calculated and recorded as the physical delay time. In this embodiment, the measured value is 1.5s. Since the sampling frequency is 20Hz, the physical delay time is converted into a sampling period, that is, it is delayed by 30 sampling periods. (4) Set up a circular buffer that can store historical data. In this embodiment, the size of the circular buffer is set to 100 floating-point spaces, which is sufficient to cover the maximum physical delay of 5 seconds. Store the air intake temperature in chronological order and perform the following operations at each sampling moment: store the air intake temperature collected at that moment into the head of the circular buffer. As new data continues to flow in, the old data will move to the depth of the queue in sequence. For the return air temperature collected at that moment, find the air intake temperature 30 sampling periods ago and form a data pair with the return air temperature collected at that moment. (5) The inlet and outlet air temperatures collected at each moment are processed by spatiotemporal alignment to obtain data pairs. As the production process continues, these data pairs, which are arranged in chronological order and generated continuously, form a set and are named the real-time temperature sequence.

[0028] After spatiotemporal alignment, for time... The intake air temperature is denoted as The corresponding return air temperature is denoted as .

[0029] It should be noted that the specific hardware selection and parameter settings described above are merely preferred embodiments of the present invention and should not be construed as limiting the scope of protection of the present invention. For example, the central processing unit can be replaced by a programmable logic controller, digital signal processor, field-programmable gate array, edge computing gateway, or other computing devices with logical operation capabilities; the acquisition unit can be replaced by other forms of sensing devices such as thermocouples, infrared non-contact sensors, and MEMS integrated sensors; the spatiotemporal alignment of data is not limited to a circular buffer and can employ delayed linked lists, cross-correlation calibration, real-time compensation algorithms based on fluid dynamics models, etc.; any modifications, equivalent substitutions, or improvements made based on the technical concept of physical spatiotemporal matching of inlet and outlet air related data are equivalent technical means of the present invention and should be included within the scope of protection of the present invention.

[0030] S2: Calculation of process progress factor.

[0031] Specifically, the inlet and outlet air temperature difference is calculated based on the inlet and outlet air temperatures. The inlet and outlet air temperature difference is then integrated over time to obtain the cumulative heat absorption of the dough at the current moment. The cumulative heat absorption of the dough is then input into a preset S-shaped growth model for nonlinear mapping to obtain a process progress factor, which is used to evaluate the physical degree of dough transformation from a raw state to a mature state, and to provide an evaluation index for the dynamic adjustment of subsequent control strategies.

[0032] Inlet air temperature The temperature of the high-temperature heat-carrying gas flow before entering the furnace reflects the level of thermal potential energy supplied by the heat source; while the return air temperature... This represents the temperature of the return gas after it has flowed over the surface of the dough and completed heat exchange, reflecting the residual state after the release of thermal energy; the difference between the two is defined as the inlet and outlet air temperature difference, denoted as . This represents the dough at a certain moment. Instantaneous heat exchange intensity, The higher the value, the more heat is released when the hot airflow passes through the dough. The smaller the value, the closer the center temperature of the dough is to the airflow temperature, and the heat exchange process tends to be in equilibrium.

[0033] Since baking is essentially a process in which a substance absorbs heat and undergoes complex physicochemical changes, such as starch gelatinization, protein denaturation, and crust caramelization, the change in the temperature difference between the inlet and outlet air at a single moment can only determine the heat absorption rate of the dough at that moment, and cannot characterize the overall ripening progress of the dough. In addition, most burners in baking sites typically use an on-off control strategy, resulting in a sawtooth-shaped periodic fluctuation in temperature. If control is based solely on the instantaneous temperature difference, it is very easy to cause oscillations in the control system. Therefore, this embodiment uses the trapezoidal numerical integration method to calculate the cumulative heat absorption of the dough, as shown in the following formula:

[0034] in: For a moment The dough accumulates heat; For a moment The dough accumulates heat; For a moment The temperature difference between the incoming and outgoing air; For a moment The temperature difference between the incoming and outgoing air; The sampling period.

[0035] At the moment the baking program starts, that is At any given time, a double-precision floating-point variable is set in the central processing unit as an accumulated heat register, and its initial value is cleared to zero; The central processing unit reads the current temperature difference between the inlet and outlet air and the temperature difference between the inlet and outlet air at the previous moment. According to the principle of calculus, the small area enclosed by these two points in time and the time axis is approximately a trapezoid. The area of ​​the trapezoid is calculated, that is, the arithmetic mean of the two temperature differences between the inlet and outlet air is calculated and then multiplied by the sampling period to obtain the heat increment of the dough at the current moment. The calculated heat increment of the dough at the current moment is added to the cumulative heat register, and the result is updated to the cumulative heat absorbed by the dough at the current moment.

[0036] It can be seen that, through the above integral calculation, the result is... It is a physical quantity that monotonically increases with the baking process, characterizing the dough from the start of heating to the end of the baking process. The total effective heat obtained reflects the degree of cooking inside the dough; and because integral operations have a natural low-pass filtering characteristic in the frequency domain, when the temperature fluctuates at high frequencies due to the start and stop of the burner, i.e. The value of fluctuates greatly, and these positive and negative fluctuations cancel each other out during the integration process, thus affecting the output. The curve showing the change over time is a smooth, rising curve, which allows the control system to ignore local temperature disturbances and always track the macroscopic trend of dough maturation.

[0037] It should be noted that the trapezoidal integration method used in this embodiment is only a preferred implementation of numerical integration, which aims to balance the computational accuracy and the processor's computational load. Those skilled in the art will understand that, if computational resources permit, the more accurate Simpson integration method or Runge-Kutta method can also be used. In low-cost applications where high accuracy is not required, it can be simplified to the rectangular integration method. All technical means that use the time-domain accumulation of the temperature difference between the inlet and outlet air to characterize the total heat are within the protection scope of this invention.

[0038] Although It reflects the heat accumulation of the dough, but its value range varies with the baking time and the total weight of the dough in a single baking, making it difficult to directly use as the feedback gain of the control system. Therefore, this implementation introduces the logistic function model as a mapping tool to perform mapping processing. The physical basis for choosing this model is that the maturation process of dough is not linear, but exhibits the characteristics of a distinct biochemical S-shaped growth curve, specifically divided into three stages: Latent period: In the early stage of baking, that is, the expansion period of the dough, the heat is mainly used to raise the temperature of the dough, starch gelatinization has not yet started on a large scale, and the physical shape changes slowly. Logarithmic growth phase: The transitional period in baking, when the center temperature of the dough reaches the gelatinization point, the volume expands rapidly, and the internal pore structure is quickly reorganized. This is the period of most dramatic physical changes. Stabilization period: The later stage of baking, that is, the shaping period of the dough. The dough skeleton is formed, the rate of moisture evaporation decreases, and the browning and caramelization are mainly based on the Maillard reaction of the crust. The physical structure tends to be stable.

[0039] Based on the above three periods, a nonlinear mapping relationship is constructed to calculate the process progress factor characterizing dough maturity. The specific relationship is as follows:

[0040] in, For a moment The process progress factor has a value range of (0,1); For a moment The dough accumulates heat; The growth rate parameter reflects the length of the dough's transition period. The larger the value, the faster the dough transitions from the expansion stage to the shaping stage, which is mostly seen in small dinner rolls. The smaller the value, the smoother the transition, which is mostly seen in large European-style breads. This is the heat threshold parameter, corresponding to the critical point where the dough completes its volume expansion and begins to set, and corresponding to the central inflection point of the S-shaped growth curve. Less than hour, The value increases slowly, but once it exceeds... , The value will jump rapidly; is the base of the natural logarithm.

[0041] when When it approaches 0, the determination time is... The dough is in the expansion stage. At this time, the dough skin has not yet formed and is extremely sensitive to environmental humidity. It is necessary to suppress the wind speed to prevent the skin from forming. When it approaches 1, the determination time is... The dough is in the setting stage. At this time, the surface of the dough has hardened and set, and it is no longer afraid of drying out. The control system allows for increased convection heat transfer by increasing the fan speed to obtain a uniform caramel color.

[0042] It can be seen that the process factor obtained through this nonlinear mapping calculation allows the control system to no longer mechanically rely on time to adjust the wind speed, but to adaptively make smooth and precise adjustments based on the actual heat absorbed by the dough.

[0043] It should be noted that the growth rate parameter and heat threshold parameters These are the core variables that determine whether the control strategy can accurately match the dough ripening rhythm. To ensure the physical authenticity of these two parameters, rather than relying on subjective experience, this embodiment comprehensively calibrates them through a standard baking experiment. The specific steps are as follows: 1. Constructing a standard thermodynamic experimental environment In this embodiment, a standard French baguette was selected as the experimental target, and the total weight of the dough sample loaded in a single batch was set to 50 kg. The hot air rotary oven was placed in the standard open-loop operation mode: the heating temperature was set to a constant 230℃, the circulating fan frequency was set to a constant 45Hz, and all automatic control logic was turned off to ensure the stability of external input variables during the experiment, so as to simply examine the thermal response characteristics of the dough itself. 2. Deploy a dual-modal non-contact monitoring system Two sets of high-frequency monitoring equipment were installed outside the high-temperature observation window in the furnace: Volume monitoring: A laser displacement sensor is used to record the vertical height H of the dough sample relative to the bottom of the baking pan at a frequency of 1Hz. This indicator represents the degree of gas expansion inside the dough. Colorimetric monitoring: A non-contact online colorimeter is used to focus on the central area of ​​the dough sample surface and record the luminance value L in the CIELAB color space at a frequency of 1Hz. This index characterizes the degree of caramelization reaction on the dough surface, that is, the degree of Maillard reaction. The lower the luminance value, the darker the color.

[0044] 3. Extract characteristic inflection points of physical phase transitions Calculate the first derivative of the curve of H over time. Since the dough expands continuously in the early stages of baking and its volume remains essentially constant after setting, the moment the derivative first converges to 0 indicates that the dough's framework has hardened, its volume has stopped increasing, and its internal pore structure has fully set. Record this moment as 0. ; To find the second derivative of L over time, since the dough undergoes browning and caramelization during the shaping period, and the initial stage of this process corresponds to a sharp increase in the rate of decrease of L over time from a stable state, this implies that the second derivative has a negative extremum. Therefore, by iterating through the curve of the second derivative, we find the time corresponding to the minimum point and record that time as [value missing]. .

[0045] 4. Parameter Inversion and Consolidation Calculate the time separately With time The cumulative heat absorbed by the dough sample is denoted as . and ,calculate and The arithmetic mean of these parameters is the caloric threshold parameter. ; To ensure that the state transition process of the control system strictly falls within the physical phase transition range of the dough, this embodiment utilizes the effective response range characteristics of the logistic function to adjust the growth rate parameter. Conversely, for the linear principal domain of the S-shaped growth curve, i.e., the interval where the function value rises from 0.1 to 0.9, covering 80% of the state changes, this linear principal domain corresponds to the transition period of dough baking. According to the mathematical definition of the logistic function, the span of the independent variable required for the function value to change from 0.1 to 0.9 is... Due to time With time The intermediate time corresponds to the aforementioned transition period, therefore Thus, the growth rate parameter is obtained. .

[0046] It should be noted that the specific implementation of the calculation of the process progress factor in the above embodiments is only a preferred embodiment of the technical concept of the present invention; those skilled in the art should understand that the calculation of the cumulative heat absorption of dough is not limited to the trapezoidal integral method, and the same effect can be achieved by using the Simpson integral method, the Runge-Kutta method, and the heat load estimation model based on neural networks; the mapping model of the process progress factor is not limited to the logistic function, and the use of the hyperbolic tangent function, the Gaussian error function, the piecewise linear function, and the membership function based on fuzzy logic are all equivalent technical means of the present invention; in addition, the growth rate parameter and heat threshold parameters The monitoring methods used in the calibration experiment can be replaced according to actual conditions. For example, machine vision can be used to extract the outline area of ​​the dough to replace laser ranging, or an infrared thermal imager can be used to monitor the surface temperature distribution to replace a colorimeter. Any modifications, equivalent substitutions, or improvements made based on the core technical idea of ​​using heat accumulation rather than absolute time to characterize the baking stage and adjusting the control strategy accordingly should be included within the scope of protection of this invention.

[0047] S3: Humidity sensitivity coefficient calculation.

[0048] Specifically, based on real-time humidity sequences and process factors, a humidity sensitivity coefficient that regulates wind speed is calculated.

[0049] Since different types of dough, such as French hard bread, Japanese soft toast, and Danish pastry, have different sensitivities and tolerances to environmental humidity, in order to achieve differentiated control, it is necessary to establish a digital recipe library in advance based on the dough recipe and the equipment manual of the hot air rotary oven. Before baking begins, the central processing unit calls the optimal process humidity and humidity deviation tolerance threshold of the dough from the digital recipe library according to the dough to be processed.

[0050] To achieve the control objectives of retaining moisture according to the dough recipe in the early stages of baking and removing wind speed restrictions in the later stages of baking, this embodiment constructs a nonlinear decay model driven by process progress factors to calculate the humidity sensitivity coefficient. The specific relationship is as follows:

[0051] in, For a moment The humidity sensitivity coefficient; For a moment Process progress factors; The humidity deviation is obtained based on real-time humidity and humidity deviation tolerance threshold, specifically... , For optimal process humidity, For a moment Real-time humidity; This refers to the humidity deviation tolerance threshold. is the base of the natural logarithm.

[0052] In the early stages of baking, Approaching 0, The term approaches 1, and the control system follows accordingly. and The humidity sensitivity coefficient is adjusted based on the deviation, thereby limiting the wind speed to ensure the dough is in a humid environment; in the later stages of baking, Approaching 1, When the value approaches 0, regardless of the humidity deviation, the humidity sensitivity coefficient approaches 1. The control system ignores the humidity limit and allows the wind speed to be increased to improve the heat exchange efficiency.

[0053] It should be noted that the nonlinear decay model based on process progress factors constructed in this embodiment is only a preferred mathematical form for implementing time-varying constraint logic; those skilled in the art should understand that the calculation of the humidity sensitivity coefficient is not limited to the above form, and piecewise linear interpolation, polynomial fitting, and dynamic weights based on fuzzy logic can also be used to achieve the same function; at the same time, the implementation of the digital recipe library is not limited to local storage, but can also be implemented based on cloud databases, edge computing gateways, etc. of the Industrial Internet of Things; any technical solution that uses variables characterizing the physical stage of baking to dynamically adjust the sensitivity of the control system to environmental parameters, such as humidity, pressure, and oxygen content, and introduces a recipe differentiation threshold accordingly should be included within the protection scope of this invention.

[0054] S4: Target wind speed command generation.

[0055] Specifically, the appropriate sensitivity coefficient is converted into an execution command to drive the circulating fan, thereby enabling real-time control of the air speed in the hot air rotary furnace.

[0056] The central processing unit reads the corresponding base fan speed of the dough to be processed from the digital recipe library. Utilizing the temperature difference between the inlet and outlet air A positive gain compensation term is constructed to obtain the intermediate wind speed value, which is then combined with the humidity sensitivity coefficient. The target wind speed command is obtained, and the specific relationship is as follows:

[0057] in, For a moment The target wind speed command; Base wind speed; For a moment The temperature difference between the incoming and outgoing air; The rated full-load temperature difference benchmark is obtained from a digital recipe library and represents the standard temperature difference generated when the dough in the hot air rotary oven reaches full load and the burner is operating at full power. Humidity sensitivity coefficient; The term is a positive gain compensation term; The item represents the intermediate wind speed value.

[0058] In the early stages of baking, process factors Approaching 0, humidity sensitivity coefficient The value is low, even though at this time Approaching or even exceeding However, due to The value is low, and the final calculated value is... It's still very small, so the control system reduces the fan speed to prevent a crust from forming on the dough surface; in the later stages of baking, When the value approaches 1, the control system removes the humidity restriction, allowing the wind speed to be increased to accelerate the Maillard reaction on the dough surface.

[0059] Uncontrollable factors exist at the processing site, such as steam pipe ruptures, accidental opening of the furnace door, or sensor probe malfunctions. These faults typically manifest as non-physical abrupt changes in humidity readings. To prevent the control algorithm from outputting dangerous commands based on erroneous data, such as blowing air at full speed when the furnace door is open, causing heat backflow and scalding the operator, this embodiment introduces an abnormal threshold to limit the target wind speed command, as follows: For time Real-time humidity, calculate its rate of change. ,in For a moment Real-time humidity, For a moment Real-time humidity, The sampling period; Preset abnormal threshold In this embodiment, 10%RH / s is used. At that time, maintain the calculated Unchanged; when At that time, the target wind speed command will be locked at a safe idle speed value until... The safe idle speed value represents the minimum airflow frequency required to maintain micro-airflow circulation inside the furnace to prevent the heat exchanger from overheating and being damaged due to heat accumulation when heating is stopped. It is obtained from a digital recipe library.

[0060] The control system utilizes a built-in serial communication interface; this embodiment employs the RS-485 physical layer to transmit floating-point data. The communication frame is encapsulated as a standard Modbus-RTU communication frame and sent to the frequency converter that controls the circulating fan, thereby adjusting the output voltage frequency to change the speed of the circulating fan.

[0061] Figure 2 The figure shows the adaptive control curve of the entire baking process provided by the embodiment of the present invention. During the expansion period, the process progress factor approaches 0 and the target wind speed command is low. During the transition period, the process progress factor shows a rapid upward trend typical of the logistic function, and the target wind speed command accelerates smoothly during this period. During the browning and caramelization period, the process progress factor stabilizes at 1, indicating that the dough has been fully formed. The target wind speed command is continuously adjusted according to the change of the temperature difference between the inlet and outlet air, promoting the Maillard reaction on the surface of the dough.

Claims

1. A method for real-time control of air velocity in a hot air rotary furnace based on sensor data, characterized in that, include: The temperature and humidity data of the heating chamber and furnace are collected in real time by a preset data acquisition unit and preprocessed to obtain a real-time temperature sequence and a real-time humidity sequence. The real-time temperature sequence includes the inlet air temperature and return air temperature of the heating chamber, and the real-time humidity sequence includes the real-time humidity inside the furnace. The difference between the inlet air temperature and the return air temperature is calculated to obtain the inlet and outlet air temperature difference. The inlet and outlet air temperature difference is then integrated over time, and the integration result is mapped through a preset nonlinear mapping model to obtain the process progress factor. The difference between the real-time humidity and the preset reference value is calculated to obtain the humidity deviation. Based on the humidity deviation and the process progress factor, the humidity sensitivity coefficient is calculated. The preset base wind speed is corrected using the humidity sensitivity coefficient to generate a target wind speed command. The target wind speed command is then used to drive the circulating fan, thereby achieving real-time control of the wind speed in the hot air rotary furnace.

2. The real-time wind speed control method for a hot air rotary furnace based on sensor data according to claim 1, characterized in that, The preset nonlinear mapping model is specifically an S-shaped growth model. The time integration operation specifically includes: cumulatively calculating the change of the inlet and outlet air temperature difference over time to obtain the cumulative heat absorption of the dough; inputting the cumulative heat absorption of the dough as an independent variable into a preset logistic function model for nonlinear mapping, calculating a value between 0 and 1, which is defined as the process progress factor; wherein, the growth center of the logistic function model is determined by a preset heat threshold parameter, and the process progress factor increases with the increase of the cumulative heat absorption of the dough.

3. The real-time wind speed control method for a hot air rotary furnace based on sensor data according to claim 1, characterized in that, The humidity sensitivity coefficient includes: constructing an attenuation weight term using the process progress factor to dynamically adjust the humidity deviation, thereby obtaining the humidity sensitivity coefficient; wherein, when the process progress factor approaches 0, the humidity sensitivity coefficient decreases as the humidity deviation increases; when the process progress factor approaches 1, the humidity sensitivity coefficient approaches 1.

4. The real-time wind speed control method for a hot blast rotary furnace based on sensor data according to claim 1, characterized in that, The target wind speed command includes: obtaining a preset base wind speed, constructing a positive gain compensation term using the inlet and outlet air temperature difference, multiplying the base wind speed by the positive gain compensation term to obtain an intermediate wind speed value; and multiplying the intermediate wind speed value by the humidity sensitivity coefficient to obtain the target wind speed command.

5. The real-time wind speed control method for a hot air rotary furnace based on sensor data according to claim 1, characterized in that, The preprocessing includes: denoising the collected temperature data to filter out noise, and simultaneously performing spatiotemporal alignment on the temperature data to obtain the real-time temperature sequence.

6. The real-time wind speed control method for a hot blast rotary furnace based on sensor data according to claim 1, characterized in that, The inlet air temperature is collected using a first temperature sensor located behind the outlet guide plate of the heating chamber, and the return air temperature is collected using a second temperature sensor located at the center of the return air inlet grille of the furnace.

7. The real-time wind speed control method for a hot air rotary furnace based on sensor data according to claim 1, characterized in that, Specifically, the driving mechanism for the circulating fan involves converting the target wind speed command into a corresponding frequency control signal, sending the frequency control signal to the frequency converter controlling the circulating fan via a serial communication interface, and adjusting the output voltage frequency of the frequency converter according to the frequency control signal, thereby changing the rotational speed of the circulating fan.

8. The real-time wind speed control method for a hot blast rotary furnace based on sensor data according to claim 2, characterized in that, The calorie threshold parameter was determined through a comprehensive standard baking experiment.

9. The real-time wind speed control method for a hot blast rotary furnace based on sensor data according to claim 1, characterized in that, The method further includes: monitoring the rate of change of the real-time humidity in real time, and when the rate of change exceeds a preset abnormal threshold, locking the target wind speed command to a preset safe idle speed value until the rate of change is lower than the abnormal threshold.

10. A real-time wind speed control system for a hot air rotary furnace based on sensor data, characterized in that, include: A processor and a memory, wherein the memory stores computer program instructions that, when executed by the processor, implement the real-time wind speed control method for a hot air rotary furnace based on sensor data according to any one of claims 1-9.

Images