Tunnel furnace energy consumption optimization automatic regulation and control method

Abstract

The invention provides a tunnel furnace energy consumption optimization automatic regulation and control method, which comprises the following steps: acquiring a reduced temperature set value of a temperature zone at the discharge end of a conveyor belt and an overall temperature distribution state in a tunnel furnace, and evaluating the total adjustment amount of the temperature set value of each zone to obtain an optimized temperature set value; according to the optimized temperature set value and the continuous monitoring data of the conveyor belt speed, the real-time change trend of the thermal penetration depth of the baked product is determined, and the incidence relation of the updated thermal penetration and stay duration data is determined; according to the confirmed accumulated data of the uniform level and the running speed change record, determining the duration data of the new baked product staying in each temperature area; and identifying the stability level of the current running speed of the conveyor belt according to the new stay duration data, and obtaining a comprehensive thermal penetration depth equalization result as a regulation ending mark.

Description

Technical Field

[0001] This invention relates to the field of information technology, and in particular to an automatic control method for optimizing energy consumption in tunnel furnaces. Background Technology

[0002] Tunnel oven baking is a crucial step in food processing, widely used in the large-scale production of bread, biscuits, and other products. Its core principle lies in ensuring uniform baking both internally and externally through precise temperature and time control, while simultaneously optimizing energy consumption to improve production efficiency. However, with increasing consumer demands for product quality and rising energy costs, achieving efficient and uniform baking in dynamic production environments while reducing energy consumption has become a critical challenge for the industry. Existing baking methods typically rely on fixed temperature settings and conveyor belt speeds, controlling the baking process through preset process parameters. However, these methods often exhibit limitations when faced with unavoidable variables during production. For example, to adapt to order demands or equipment status, the conveyor belt speed may need temporary adjustments, and fixed parameters cannot flexibly handle this, leading to uneven heating of the product. Especially when speed changes frequently, traditional methods struggle to adapt in real-time to the actual heating conditions of the product in different temperature zones, accurately grasp the actual residence time of the product in each temperature zone of the tunnel oven, and adjust the heating strategy accordingly, thus affecting product quality and energy efficiency. Residence time directly affects the depth of heat penetration into the product and is a key factor determining baking uniformity. If the dwell time is too short, heat cannot be fully conducted to the interior of the product, potentially resulting in uncooked interior; conversely, if the dwell time is too long, the surface may be overheated, affecting taste and quality. Further complicating matters, tunnel ovens are typically divided into multiple temperature zones, and the temperature settings of each zone need to be coordinated to adapt to differences in dwell time caused by speed variations. Current technology struggles to achieve this dynamic coordination, making precise control of heat penetration depth difficult. This uneven heat penetration caused by variations in dwell time not only affects product quality but also leads to energy waste. Therefore, how to monitor the dwell time of the product in each temperature zone in real time under dynamically changing conveyor belt speeds, and dynamically adjust the temperature settings of each zone accordingly to ensure uniform heat penetration depth and optimal energy consumption, has become a critical problem that urgently needs to be solved in tunnel oven baking production. Summary of the Invention

[0003] This invention provides an automatic control method for optimizing energy consumption in a tunnel furnace, mainly comprising:

[0004] The system acquires information on the conveyor belt speed and the position of the baked products in the tunnel oven. It extracts heat penetration depth data from the temperature distribution, identifies changes in operating speed, and assesses the trend deviation value of the impact of these speed changes on heat penetration uniformity. Based on the trend deviation value and residence time data, it determines the heat conduction status of the product in each temperature zone, obtains an initial heat penetration index, and adjusts the temperature setpoint of the conveyor belt feed end temperature zone when the initial heat penetration index is below the standard value. It then extracts adjustment amplitude data from the adjusted temperature setpoint, assesses the heating risk in the conveyor belt discharge end temperature zone, and adjusts the temperature setpoint of the discharge end temperature zone accordingly. The adjusted temperature setpoints of the discharge end temperature zone are obtained, and the total adjustment amount of the temperature setpoints of each zone is evaluated to generate a temperature configuration combination. Based on the energy consumption control scheme and conveyor belt speed monitoring data, the correlation between heat penetration and residence time is determined. The heat penetration trend is extracted from the correlation, and when it matches the standard curing requirements, the temperature setpoints of all temperature zones are obtained to determine that the internal temperature distribution of the product has reached a uniform level. After determining that the internal temperature distribution of the product has reached a uniform level, new residence time data is generated. Based on the new residence time data, the stability of the conveyor belt running speed is identified, and a heat penetration depth equalization result is generated.

[0005] Furthermore, the acquisition of the tunnel oven conveyor belt running speed and baking product position information, extraction of heat penetration depth data from the temperature distribution state, identification of running speed changes, and evaluation of the trend deviation value of the impact of running speed changes on heat penetration uniformity include:

[0006] The system acquires real-time conveyor belt speed data and the position coordinates of the baked product inside the oven, collects surface temperature distribution data of the product, calculates the heat penetration depth from the product surface to the interior, and records the amplitude and frequency of the conveyor belt speed change within a time window. Based on the amplitude and frequency of the change, the system calculates the dwell time of the product in each temperature zone, extracts temperature values ​​at different depths of the product from the temperature distribution data, calculates the temperature gradient distribution, compares it with a standard baking curve to determine the degree of deviation of the heat penetration depth, and generates an influence coefficient. The system then multiplies the influence coefficient by the heat received by the product to generate a heat penetration adjustment value for each zone, predicts the trend of heat penetration depth change at the next moment, and calculates the trend deviation value.

[0007] Furthermore, the step of determining the heat conduction status of the product in each temperature zone based on the influence trend deviation value and residence time data, obtaining the initial heat penetration index, and adjusting the temperature setpoint of the conveyor belt feed end temperature zone when the initial heat penetration index is lower than the standard value includes:

[0008] Based on the aforementioned influence trend deviation value and residence time data, the heat received by the product in each temperature zone is calculated, the heat conduction rate is determined, and heat conduction state distribution data is generated. The initial temperature value of the product is extracted from the product location information, the temperature zones traversed by the product and the duration of heating are traced, and the initial heat penetration index is calculated. The initial heat penetration index is compared with the standard heat penetration curve to determine the temperature adjustment range and generate the corrected temperature setpoint for the feed end temperature zone. The feed end heating power configuration is updated according to the corrected temperature setpoint, the temperature increment is allocated, and the adjusted feed end temperature zone temperature setpoint is generated.

[0009] Furthermore, the step of extracting adjustment range data from the adjusted temperature setpoint, assessing the heating risk in the temperature zone at the conveyor belt outlet, and adjusting the temperature setpoint in the outlet temperature zone includes:

[0010] The difference between the current temperature value and the original temperature value is extracted from the adjusted temperature setpoint of the feed end temperature zone to generate adjustment amplitude data, calculate the additional heat accumulation value, compare it with the safe heat threshold, and determine the heating risk level; the temperature reduction coefficient is determined according to the risk level, the temperature reduction amount at the discharge end is calculated, and it is allocated to the discharge end heating unit; the discharge end heating control parameters are updated according to the temperature reduction amount, the heat supply balance is verified, and the adjusted temperature setpoint of the discharge end temperature zone is generated.

[0011] Furthermore, the process of obtaining the adjusted temperature setpoint of the discharge end temperature zone, evaluating the total adjustment of the temperature setpoint of each zone, and generating a temperature configuration combination includes:

[0012] The adjusted temperature setpoint of the discharge end temperature zone and the real-time readings of each temperature sensor are obtained. The total adjustment amount is obtained by accumulating the absolute values ​​of the adjustment amplitudes of all zones. The power consumption value corresponding to each temperature is queried from the energy consumption records established from historical production data. The temperature setpoint of each zone is iteratively optimized using a genetic algorithm with the goal of minimizing total energy consumption and the constraint of ensuring that the product center temperature reaches the preset ripening temperature. The temperature configuration combination with the minimum energy consumption is searched.

[0013] Furthermore, determining the correlation between heat penetration and residence time based on the temperature configuration combination and conveyor belt speed monitoring data includes:

[0014] Based on the temperature configuration combination and conveyor belt speed monitoring data, the real-time position coordinates of the product are obtained, the current heat penetration depth value is calculated, a depth value sequence is generated, and the changing trend is identified; the changing trend is compared with the ideal heat penetration trajectory, the deviation value is calculated, and the correlation between heat penetration depth and residence time is fitted.

[0015] Furthermore, the step of extracting the heat penetration trend from the correlation and matching it with the standard curing requirements, obtaining the temperature setpoints for all temperature zones, and determining that the internal temperature distribution of the product has reached a uniform level includes:

[0016] Extract the slope change sequence from the correlation to generate a continuous curve of heat penetration depth changing with time and identify trend characteristics; compare the trend characteristics with the aging temperature standard to determine the matching status, obtain the temperature setpoint and real-time temperature readings of all temperature areas; calculate the temperature values ​​of each layer of the product based on the real-time temperature readings to confirm that the internal temperature distribution of the product has reached a uniform level.

[0017] Furthermore, after determining that the internal temperature distribution of the product has reached a uniform level, new residence time data is generated, including:

[0018] Based on the uniformity level and operating speed change data, query historical energy consumption records, identify the parameter combination with the lowest energy consumption, and determine the direction of optimization control; based on the optimization control direction, obtain product location information and temperature distribution status, establish temperature value mapping, and calculate heat accumulation trajectory; based on the heat accumulation trajectory and conveyor belt speed data, calculate the dwell time of the product in each temperature zone, and generate new dwell time data.

[0019] Furthermore, the step of identifying the conveyor belt speed stability based on the new residence time data and generating a heat penetration depth equalization result includes:

[0020] The dwell time values ​​of each region are extracted from the new dwell time data, the conveyor belt speed sequence is obtained, the coefficient of variation is calculated, the stability of the running speed is judged, the thermal penetration value deviation is calculated, and the thermal penetration depth equilibrium result is generated.

[0021] The technical solutions provided by the embodiments of the present invention may include the following beneficial effects:

[0022] This invention discloses an automatic energy consumption optimization and control method for tunnel ovens. Addressing the problems of uneven heat penetration and excessive energy consumption caused by variations in conveyor belt speed during baking, the method acquires real-time data on conveyor belt speed, product position, and temperature distribution to extract a heat penetration depth index and identify the impact of speed changes on heat penetration uniformity. When the initial heat penetration index is lower than the standard value, the temperature setpoint at the feed end is dynamically increased, while the potential overheating risk at the discharge end is assessed, and its temperature setpoint is appropriately reduced to ensure uniform overall temperature distribution. Based on the optimized temperature setpoint and continuously monitored operating speed data, this invention refines the correlation between heat penetration trend and residence time through a feedback adjustment mechanism. After matching standard ripening requirements, the uniformity of internal temperature distribution in the product is verified. Finally, by combining residence time and speed stability analysis, an energy consumption optimization control scheme is generated. This invention significantly improves the heat penetration uniformity of baked products, reduces energy consumption, and achieves efficient and stable baking process control. Attached Figure Description

[0023] Figure 1 This is a flowchart of an automatic control method for optimizing energy consumption in a tunnel furnace according to the present invention. Detailed Implementation

[0024] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0025] like Figure 1 This embodiment of the automatic control method for optimizing energy consumption in a tunnel furnace may specifically include:

[0026] Step S101: Obtain the current running speed of the tunnel oven conveyor belt and the position information of the baked products, extract the heat penetration depth data from the temperature distribution state, and evaluate the trend deviation value of the influence of the running speed change on the heat penetration uniformity of the baked products.

[0027] Real-time speed data of the conveyor belt in the tunnel oven and the position coordinates of the baked products inside the oven are acquired. Infrared thermal imagers are used to collect images of the product surface temperature distribution. The penetration depth of heat from the product surface to the interior is calculated using the heat conduction equation Q = kA(T1-T2)d / L, where k is thermal conductivity, A is the heat transfer area, T1 and T2 are the surface and interior temperatures respectively, d is the penetration depth, and L is the product thickness. The amplitude and frequency of the conveyor belt speed change within a preset time window are recorded. The actual residence time of the product in each temperature zone is calculated based on the amplitude and frequency of the change. Temperature values ​​at different depths on the product cross-section are extracted from the thermal imaging data. The temperature gradient distribution is obtained by calculating the temperature difference between adjacent depths. The deviation of the heat penetration depth is judged by comparing the temperature gradient at the corresponding moment in the standard baking curve, and the influence coefficient of speed change on heat penetration uniformity is obtained. This influence coefficient is multiplied by the cumulative heat received by the product in different regions to obtain the heat penetration adjustment value for each region. Linear regression analysis is used to predict the trend of heat penetration depth change at the next moment based on historical heat penetration depth data. The initial value of the trend deviation is calculated based on the difference between the predicted value and the target heat penetration depth. By comparing the initial value of the trend deviation with the heat penetration adjustment value of each region, the temperature region where the deviation exceeds the preset threshold is identified as the region with uneven heat penetration. The uniformity level is evaluated by calculating the standard deviation of the temperature values ​​at each measurement point inside the product, and the trend deviation value of the influence of the change in running speed on the heat penetration uniformity of the baked product is obtained.

[0028] Specifically, in one embodiment, real-time monitoring of the tunnel oven conveyor belt speed and product position is achieved through a combined sensor array. A high-precision rotary encoder is installed on the conveyor belt drive shaft, outputting a pulse signal every certain rotation angle. The control system calculates the real-time linear speed of the conveyor belt based on the pulse frequency. Simultaneously, a photoelectric sensor is installed at the tunnel oven entrance. When baked products pass through, a timer is triggered, and the product's position coordinates at any given moment inside the oven are calculated by combining the conveyor belt speed data. An infrared thermal imager is installed in the tunnel oven's observation window, with a sampling frequency set to 10 frames per second to acquire images of the temperature distribution on the product surface.

[0029] Specifically, the application of the heat conduction equation Q = kA(T1-T2)d / L requires accurate measurement of various parameter values. Thermal conductivity k is pre-determined based on product type; for bread products, it is typically in the range of 0.05 to 0.15 W / (m·K), while for biscuits it is between 0.08 and 0.20 W / (m·K). The heat transfer area A is calculated by extracting the product outline from the thermal imaging image using an image recognition algorithm. The surface temperature T1 is read directly from the thermal imaging data, while the internal temperature T2 is measured by a miniature temperature probe embedded in the product center or estimated based on a temperature field model established from historical baking data. The product thickness L is measured and recorded at the production line entrance using a laser rangefinder. The penetration depth d, calculated from these parameters, reflects the actual depth of heat transfer into the product, a value directly related to the product's degree of ripeness.

[0030] It's important to note that monitoring conveyor belt speed changes involves two key indicators: the magnitude of the change and the frequency of the change. The magnitude of the change is calculated by the difference in speed within adjacent time windows; a change is recorded as valid when the difference exceeds a preset threshold. The frequency of the change is the number of valid changes per unit time. These two indicators together reflect the stability of the production line operation and are crucial for subsequent heat penetration analysis.

[0031] In one possible implementation, the dwell time is calculated using a piecewise integration method. The tunnel furnace is divided into multiple zones based on a set temperature, and the length of each zone is known. When the conveyor belt speed is constant, the dwell time equals the zone length divided by the conveyor belt speed. However, in actual production, the speed frequently changes, so it is necessary to track the product position in real time and record the times when it enters and leaves each zone; the difference between these two times is the actual dwell time. The temperature gradient is extracted by radially scanning the thermal imaging data. Temperature values ​​are sampled at fixed intervals from the product surface towards the center, and the temperature difference between adjacent sampling points divided by the distance yields the temperature gradient at that location.

[0032] For example, the standard baking curve is established based on a large amount of historical production data. Temperature distribution data for different product types and specifications under optimal baking conditions are collected, and the standard temperature gradient range at each time point is obtained through statistical analysis. In actual production, the real-time measured temperature gradient is compared with the standard value, and the degree of deviation is determined by calculating the relative error between the two. When the relative error exceeds 15%, it is considered a significant deviation, and process parameters need to be adjusted. The calculation of the influence coefficient comprehensively considers the degree of deviation and the duration; the greater the degree of deviation and the longer the duration, the higher the influence coefficient.

[0033] Preferably, the calculation of the heat penetration adjustment value incorporates the concept of a cumulative effect. The heat generated when the product remains in a high-temperature zone continues to be conducted inwards to subsequent zones; therefore, the heat penetration of each zone is not independent but rather mutually influential. By establishing a spatiotemporal model of heat conduction, the propagation path of heat within the product is tracked, and the contribution weight of each zone to the final heat penetration depth is calculated. Linear regression analysis employs the least squares method, using the heat penetration depth at historical moments as the independent variable and the measured value at the next moment as the dependent variable, to fit and obtain a prediction model. The model input includes multiple parameters such as the current heat penetration depth, conveyor belt speed, and temperature setpoints for each zone.

[0034] For example, if the predicted value shows that the product's center temperature will be 5°C lower than the standard value when it comes out of the oven, it is determined that there is a risk of underbaking. The initial trend deviation value calculated at this time is negative, indicating that more heat input is needed. By comparing the heat penetration adjustment values ​​of each region, areas with insufficient contribution are identified. Typically, the inlet area has a lower temperature, a larger temperature difference between the inside and outside of the product when it first enters, and high heat transfer efficiency, making it suitable as a key area for compensation adjustment.

[0035] In one embodiment, a dynamic threshold method is used to identify areas of uneven heat penetration. The preset threshold is not a fixed value, but is dynamically adjusted according to the product type, specifications, and current baking progress. In the early stages of baking, the temperature difference between the inside and outside of the product is large, and the allowable deviation range is relatively wide; in the later stages of baking, the product is close to being fully cooked, and the tolerance for deviation decreases. The standard deviation calculation covers multiple measurement points on the cross-section of the product, and the measurement points are evenly distributed in a grid pattern to ensure that the internal temperature distribution can be fully reflected.

[0036] Understandably, determining the trend deviation value requires considering multiple factors. Besides the real-time measured deviation, the trend of deviation change, its duration, and its potential impact on the final product quality must also be considered. By establishing a multi-dimensional evaluation system and assigning appropriate weights to different factors, a comprehensive trend deviation value can be calculated. This value not only reflects the current state of heat penetration but, more importantly, predicts how product quality will change without intervention.

[0037] Step S102: Based on the trend deviation value of the effect of the change in running speed on the actual heat penetration uniformity of the baked product and the residence time data, determine the actual heat conduction status of the product in each temperature zone, obtain the initial heat penetration index, and when it is determined that the initial heat penetration index is lower than the standard value, increase the temperature setting value of the temperature zone at the feed end of the conveyor belt.

[0038] A heat conduction mapping table is constructed based on the influence trend deviation value and residence time data to obtain the cumulative heat received by the product in each temperature zone. The actual heat conduction rate is determined by the ratio of heat flux density to temperature difference, and the heat conduction rate of each zone is recorded to form heat conduction state distribution data. The temperature zone number corresponding to the current position of the product is identified using the heat conduction state distribution data. The initial temperature value is extracted from the initial temperature record when the product enters the tunnel furnace. Based on the product position information, all temperature zones it has passed through and the corresponding cumulative heating time are traced. The initial heat penetration index A = Q / (M*C) is calculated by the ratio of the cumulative heat Q to the product of product mass M and specific heat capacity C. The initial heat penetration index is compared with the preset standard heat penetration curve. If the initial heat penetration index is lower than the standard value and exceeds the preset threshold, the temperature increase range is determined according to the deviation value to obtain the corrected temperature setpoint of the feed end temperature zone. The heating power configuration of the feed end temperature zone is updated using the corrected temperature setpoint. The temperature increment is distributed among each heating unit at the feed end using a stepped temperature increase method to obtain the improved temperature setpoint of the conveyor belt feed end temperature zone.

[0039] Specifically, in one implementation, the heat conduction mapping table is constructed based on multi-dimensional data fusion technology. The influence trend deviation value serves as the main input parameter, reflecting the degree of influence of velocity changes on heat penetration uniformity, while the residence time data records the actual residence time of the product in each temperature zone. The mapping table adopts a two-dimensional matrix structure, with the horizontal axis representing different temperature zone numbers and the vertical axis representing time slices. Each matrix element stores the heat conduction intensity value at the corresponding spatiotemporal location. The accumulated heat is obtained through integration, that is, within each time slice, the product of instantaneous heat flow and the time interval is accumulated and summed.

[0040] Specifically, determining the heat transfer rate requires comprehensive consideration of temperature gradient and material properties. Inside the tunnel oven, a temperature sensor array is arranged every 0.5 meters, with each array containing three measuring points (top, middle, and bottom) to collect real-time air temperature data. The product surface temperature is obtained through infrared thermography, while the internal temperature is calculated based on heat conduction theory. The ratio of heat flux density to temperature difference is such that, under steady-state heat transfer conditions, heat flux density is proportional to the temperature gradient, and the proportionality coefficient is the material's thermal conductivity. The thermal conductivity of bread products varies with moisture content; it is higher during the initial baking stage when moisture content is high, and gradually decreases as moisture evaporates. This dynamic change directly affects the accuracy of the heat transfer rate calculation.

[0041] It should be noted that the formation of thermal conductivity distribution data involves spatial interpolation. Due to the limited sensor density, the temperature values ​​between adjacent measurement points are obtained through cubic spline interpolation to ensure the continuity of the temperature field. The thermal conductivity rate in each temperature region is not uniformly distributed; the thermal conductivity rate is higher near the heating element and relatively lower further away. This spatial distribution characteristic is accurately recorded through thermal conductivity distribution data, providing a basis for subsequent temperature control.

[0042] For example, product location identification is achieved using a coded positioning method. A photoelectric trigger device is installed at the entrance of the tunnel oven. When a product passes through, the entry time is recorded and a unique number is assigned. Combining conveyor belt speed data and oven length information, the accurate position of the product inside the oven is calculated in real time. Temperature zones are pre-divided according to the oven structure, typically into a preheating zone, a main baking zone, and a post-baking zone, with each main zone further subdivided into several sub-zones. When the product's position coordinates fall within a certain zone, its temperature zone number can be determined.

[0043] In one possible implementation, the calculation of the initial heat penetration index involves the comprehensive processing of multiple parameters. The initial temperature value is extracted from the temperature record when the product first enters the tunnel oven, typically room temperature or the temperature after proofing. Product mass is obtained through a weighing device at the inlet, and specific heat capacity is determined by referring to a table based on the product formulation and component ratios. The calculation of cumulative heat requires tracing the complete thermal history of the product from its entry into the tunnel oven to the present moment, including the residence time and corresponding heating intensity in each temperature zone. The initial heat penetration index is calculated by dividing the cumulative heat by the product mass and specific heat capacity. This index reflects the actual effect of heat transfer into the product.

[0044] Preferably, the standard heat penetration curve is established based on statistical analysis of a large amount of historical production data. Ideal heat penetration depth data for similar products at different baking stages are collected, and a standard curve is obtained through curve fitting. The standard value is not fixed but changes dynamically with the baking process. Heat penetration depth increases rapidly in the early stages of baking, stabilizes in the middle stages, and slightly decreases in the later stages. When the measured initial heat penetration index is lower than the standard value at the corresponding time point, and the deviation exceeds a preset threshold, it indicates a risk of insufficient heating of the product.

[0045] For example, the temperature increase is determined using a proportional adjustment principle. The larger the deviation, the higher the required temperature increase, but there is an upper limit to prevent overheating. The feed end temperature zone, as the first high-temperature area the product comes into contact with, has a significant impact on the overall baking effect when its temperature is adjusted. The corrected temperature setpoint is obtained by adding the temperature increase to the original setpoint, while also considering the equipment's heating capacity limitations and the product's heat resistance characteristics.

[0046] In one embodiment, the heating power configuration is updated through zone control. The feed-end temperature zone contains multiple independently controlled heating units, each equipped with an independent power adjustment device. The stepped temperature increment method refers to the temperature setpoint of each heating unit increasing progressively from the inlet to the outlet, forming a temperature gradient. The distribution of temperature increments follows the principle of maximizing heat transfer efficiency, providing a larger temperature increment when the product first enters and gradually decreasing the increment as the product temperature rises.

[0047] Step S103: Extract adjustment range data from the temperature setpoint of the improved conveyor belt feed end temperature zone, assess the potential overheating risk of the conveyor belt discharge end temperature zone, and reduce the temperature setpoint of the conveyor belt discharge end temperature zone.

[0048] The difference between the current temperature value and the original temperature value is extracted from the temperature setpoint of the conveyor belt feed end temperature zone after the temperature increase, and used as the adjustment range data. This adjustment range data is multiplied by the expected dwell time of the product at the discharge end to obtain the additional heat accumulation value. The potential overheating risk level is determined by comparing this additional heat accumulation value with a preset safe heat threshold. The temperature reduction coefficient at the discharge end is determined by querying a preset risk-coefficient correspondence table using the risk level. The temperature reduction amount is obtained by multiplying the current temperature setpoint at the discharge end by the temperature reduction coefficient. This temperature reduction amount is allocated from the power configuration of each heating unit at the discharge end according to the current power ratio of each unit. The heating control parameters of the discharge end temperature zone are updated based on the temperature reduction amount. By calculating the difference between the heat increment generated by the increase in feed end temperature and the heat decrement generated by the decrease in discharge end temperature, the overall heat supply balance is verified, and the reduced temperature setpoint of the conveyor belt discharge end temperature zone is obtained.

[0049] Specifically, in one implementation, the adjustment amplitude data is extracted based on precise measurement of the temperature difference. The increased temperature setpoint of the feed zone on the conveyor belt is recorded in real time by the control system and compared with the original temperature value before adjustment; the difference is the adjustment amplitude. This amplitude reflects the magnitude of temperature increase to compensate for insufficient heat penetration. The expected residence time of the product at the discharge end is calculated based on the current conveyor belt speed and the length of the discharge end area; the additional heat accumulation value is equal to the product of the adjustment amplitude and the residence time, multiplied by the thermal conductivity coefficient.

[0050] Specifically, the assessment of potential overheating risk involves multiple judgment dimensions. Preset safe calorie thresholds are established based on product type and specifications. For bread products, the safety threshold is typically set as the critical calorie value to prevent crust charring, while for biscuits, it's set as the upper limit to prevent over-brittleness. When the accumulated excess heat approaches or exceeds the safe threshold, the system determines there is an overheating risk. The risk level is divided into three levels: low, medium, and high, each corresponding to different temperature control intensities.

[0051] It should be noted that the risk-coefficient correspondence table is a mapping relationship established based on a large amount of production data statistics. Low risk corresponds to a smaller temperature reduction coefficient, such as 0.95; medium risk corresponds to a medium coefficient, such as 0.90; and high risk corresponds to a larger coefficient, such as 0.85. The current temperature setpoint at the discharge end is multiplied by the temperature reduction coefficient to obtain the adjusted target temperature, and the difference between the two is the temperature reduction amount. This dynamic adjustment mechanism based on risk level prevents overheating of the product and avoids insufficient ripening caused by excessive cooling.

[0052] Preferably, the temperature reduction is distributed among the heating units using a power percentage method. The discharge end temperature zone typically contains multiple independently controlled heating units, and the current power of each unit is acquired in real time through a power monitoring module. The temperature reduction allocated to each unit is equal to the total temperature reduction multiplied by the proportion of that unit's power to the total power. Units with higher power undertake more cooling tasks, maintaining a balanced temperature regulation.

[0053] For example, heat balance verification is achieved by comparing the heat changes at the inlet and outlet ends. The heat increment caused by the temperature increase at the inlet end is calculated based on the temperature increase value and the heating efficiency of that area, while the heat reduction caused by the temperature decrease at the outlet end also takes into account temperature changes and heat dissipation characteristics. The difference between the two reflects the change in heat supply of the entire tunnel furnace system. When the difference is close to zero, it indicates that the system has reached a heat balance state, and the outlet end temperature setpoint obtained at this time is the optimization result.

[0054] Step S104: Obtain the reduced temperature setpoint of the conveyor belt discharge end temperature zone and the overall temperature distribution in the tunnel furnace, evaluate the total adjustment amount of the temperature setpoint of each zone, and generate a temperature configuration combination.

[0055] The temperature setpoint at the conveyor belt outlet and real-time readings from temperature sensors within the tunnel furnace are acquired after the temperature reduction. The dispersion of temperature distribution is assessed by calculating the standard deviation of temperature differences between adjacent areas. If the standard deviation is lower than a preset uniformity threshold based on product type, the overall temperature distribution within the tunnel furnace is determined to be relatively uniform. The temperature adjustment range for each area is calculated based on the difference between the current temperature setpoint and the original temperature setpoint before adjustment, under the uniform condition. The total adjustment is obtained by summing the absolute values ​​of all area adjustment ranges. The power consumption value corresponding to each temperature is retrieved from the energy consumption records established from historical production data. The energy consumption value for each area is calculated by multiplying the power consumption value by the product's dwell time in each area. A genetic algorithm is used to iteratively optimize the temperature setpoints for each area with the objective function of minimizing total energy consumption and the constraint of achieving a preset ripening temperature at the product's center. The temperature configuration combination with the minimum energy consumption is searched. The control parameters for each temperature area are updated based on the temperature configuration combination. The difference between the total energy consumption after optimization and the total energy consumption before optimization, as well as the temperature adjustment ratio for each area, are recorded to obtain an optimized energy consumption control scheme that includes temperature setpoints, power allocation, and expected energy consumption.

[0056] Specifically, in one implementation, the assessment of temperature distribution uniformity is achieved using statistical methods. The tunnel oven is functionally divided into a preheating zone, a main baking zone, and a post-baking zone. Multiple temperature sensors are installed in each zone, with the sensor spacing evenly distributed according to the oven's length. The reduced temperature setpoint at the conveyor belt outlet serves as the starting point for control, and temperature readings from each sensor are collected in real time. The temperature difference between adjacent zones is obtained by subtracting the readings from adjacent sensors, and the standard deviation calculation covers all adjacent temperature differences, reflecting the degree of dispersion in the temperature distribution.

[0057] Specifically, the setting of the uniformity threshold needs to consider product characteristics and process requirements. Bread products have high requirements for temperature uniformity, and the threshold is usually set within 5℃; for biscuit products, due to their thinner thickness, the threshold can be appropriately relaxed to 8℃. When the calculated standard deviation is lower than the corresponding threshold, the system determines that the temperature distribution has reached a uniform state. This determination method avoids the impact of single-point temperature anomalies on the overall assessment and improves the accuracy of the judgment.

[0058] It's important to note that the calculation of the temperature adjustment range involves a time dimension. The original temperature setpoint before adjustment refers to the initial setpoint before any optimization operations are performed; these values ​​are determined by the process formula and stored in the control system at the start of production. The current temperature setpoint is the real-time value after multiple adjustments. The difference between the two is the temperature adjustment range for each zone; a positive value indicates a temperature increase, and a negative value indicates a temperature decrease. The total adjustment amount is obtained by summing the absolute values ​​of the adjustment ranges for all zones, reflecting the overall control intensity.

[0059] For example, the establishment and querying of energy consumption records are based on the accumulation of long-term production data. After each production batch, the actual power consumption corresponding to each temperature setpoint is automatically recorded, forming a database of temperature-power correspondence. The database is stored according to dimensions such as product type, furnace area, and seasonal factors. During a query, the most similar historical record is matched based on the current production conditions to extract the corresponding power consumption value. If no completely matching record is found, the power value is estimated using interpolation methods.

[0060] In one possible implementation, the application of genetic algorithms requires a clear definition of each element of the optimization problem. Chromosome encoding uses real numbers, with each gene locus representing a set value for a temperature range. The population size is set to 50 to 100 individuals to ensure sufficient coverage of the search space. The fitness function is defined as the reciprocal of the total energy consumption; lower energy consumption results in higher fitness. The selection operation uses a roulette wheel approach, where individuals with higher fitness have a greater probability of being selected. The crossover operation uses arithmetic crossover, where the genes of two parent individuals are combined according to random weights to produce offspring. The mutation operation adds Gaussian noise to the gene values ​​to maintain population diversity. Constraints are handled using a penalty function method; when the product's core temperature is lower than the preset ripening temperature, a penalty term is applied to the fitness value to ensure that the search results meet quality requirements.

[0061] Preferably, the calculation of regional energy consumption values ​​requires precise time parameters. The dwell time of the product in each region is calculated in real time based on the conveyor belt speed and region length, and this dwell time directly affects the energy accumulation effect. The product of the power consumption value and the dwell time yields the energy consumption value of that region, and the sum of the energy consumption values ​​of all regions is the total energy consumption. During the iterative process of the genetic algorithm, the total energy consumption of all individuals is calculated in each generation, and the individual with the lowest energy consumption that meets the constraints is selected as the optimal solution for that generation.

[0062] For example, the constraints on the product's core temperature need to consider food safety standards. The core temperature of bread products typically needs to reach above 93℃ to ensure complete internal cooking; for biscuits, due to their low moisture content, the core temperature requirement can be lowered to 85℃. This temperature is predicted using a pre-set heat conduction model. The model input includes the temperature setpoints for each region, residence time, initial product temperature, and thermophysical parameters. When the predicted core temperature meets the requirements, the corresponding temperature configuration is considered a feasible solution. The optimized result of the temperature configuration combination contains multi-dimensional information. In addition to the temperature setpoints for each region, it may also include the corresponding power allocation ratio, expected energy consumption, and product quality prediction indicators. This information is integrated to form a complete energy consumption control scheme. Based on the scheme, the PID controller parameters for each region are updated, adjusting the power output of the heating elements. Simultaneously, the total energy consumption difference before and after optimization is recorded to evaluate the optimization effect.

[0063] Step S105: Based on the continuous monitoring data of temperature configuration combination and conveyor belt speed, determine the real-time change trend of heat penetration depth of baked products, and determine the correlation between updated heat penetration and residence time data.

[0064] Based on continuous monitoring data of temperature configuration combinations and conveyor belt speed, the real-time position coordinates of the product in each temperature zone are obtained. The current heat penetration depth is calculated using the ratio of heat flux density to temperature gradient, and a sequence of depth values ​​at consecutive time points is recorded. The arithmetic mean of data points within a time window is used to identify the real-time trend of heat penetration depth. This real-time trend is compared with the ideal heat penetration trajectory established in historical production, and the deviation value and its rate of change are calculated. If the deviation value exceeds a preset threshold, a feedback adjustment mechanism is activated, and a linear relationship function between heat penetration depth and residence time is fitted using the least squares method. The linear relationship function is used to update the stored correspondence data between heat penetration depth and residence time in the system. The difference between the updated slope and intercept parameters and the original parameters is obtained, and the control weights of each temperature zone are adjusted proportionally according to the magnitude of the difference, resulting in an optimized set of correlation parameters. The expected heat penetration depth of the product at the current speed is recalculated using this set of correlation parameters, and the optimization effect is verified by comparing it with measured values. The verification results are recorded, and the correlation between heat penetration and residence time data is updated.

[0065] Specifically, in one implementation, the energy consumption control scheme is based on a real-time monitoring and dynamic adjustment mechanism. Temperature configuration parameters include multi-dimensional information such as the set temperature for each temperature zone, the heating power allocation ratio, and the temperature change rate limit. Continuous monitoring of the conveyor belt speed is achieved through a high-precision encoder with a sampling frequency of 100 times per second, ensuring real-time capture of speed changes. Product position coordinates are accurately calculated using speed integration and position correction algorithms, and the trajectory of each product within the furnace is completely recorded.

[0066] Specifically, the ratio of heat flux density to temperature gradient reflects the fundamental law of heat conduction. Heat flux density represents the amount of heat passing through a unit area per unit time, while temperature gradient is the rate of temperature change with spatial location. The ratio of the two is the thermal conductivity of the material, a parameter that dynamically changes with temperature and moisture content. The recording interval for the heat penetration depth value sequence is set to 0.5 seconds, forming dense time series data. The size of the time window is adjusted according to the product type: a 30-second window is used for bread products, and a 20-second window is used for biscuit products. The arithmetic mean of the data points within the window can effectively filter out measurement noise and identify the true trend of change. Heat penetration depth data of similar products in different production batches are collected, and the batch with the highest product quality score is selected as a reference sample. Through curve fitting technology, discrete data points are connected into a continuous trajectory curve. The trajectory curve is divided into three stages according to the baking process: rapid penetration period, stable penetration period, and penetration slowdown period. The slope and curvature characteristics of each stage are accurately recorded as a benchmark for real-time comparison. The deviation value is obtained by calculating the difference between the measured value and the ideal value, and the rate of change reflects the speed of change of the deviation. When the deviation exceeds 10% of the product thickness or the rate of change exceeds 0.5 mm per second, it is determined that adjustment is required.

[0067] For example, the activation of the feedback adjustment mechanism involves a comprehensive evaluation of multiple judgment conditions. In addition to deviation values ​​and rates of change, factors such as the current production stage, remaining baking time, and product surface condition are also considered. The specific implementation of the least squares fitting method uses matrix operations to construct a set of observation equations, and the optimal fitting parameters are obtained by solving these equations. The linear relationship function is in the form y = ax + b, where y represents the heat penetration depth, x represents the residence time, a is the slope parameter reflecting the heat conduction rate, and b is the intercept parameter reflecting the initial heat penetration state. The fitting process uses the most recent 100 data points to ensure that the function reflects the current actual situation.

[0068] In one possible implementation, the updating of the correlation data employs a gradual adjustment strategy. The stored correspondence between heat penetration depth and residence time is organized in the form of a lookup table, which includes the correspondence under different combinations of speed and temperature. The slope and intercept calculated by the new linear relationship function are compared with the original parameters; the difference between the two reflects the degree of deviation between actual production and expectations. The slope difference is mainly caused by changes in heat transfer efficiency, while the intercept difference is related to the initial state of the product.

[0069] Preferably, the adjustment of control weights follows a graded response principle. Differences are divided into three levels: less than 5% is considered a slight deviation, 5%-15% is a moderate deviation, and more than 15% is a severe deviation. According to a preset correspondence, the corresponding weight adjustment ratios are 1:1.2:1.5. The original weight of each temperature zone is determined based on its proportion of contribution to the total heat, and the adjusted weight is obtained by multiplying the original weight by the adjustment ratio. Changes in weight directly affect the response intensity of temperature control; zones with larger weights play a dominant role in temperature regulation.

[0070] For example, the correlation parameter set contains parameter information across multiple dimensions. Besides slope and intercept, it also includes statistical indicators such as goodness of fit, confidence interval, and effective data range. These parameters together constitute a complete description of the correlation, enabling the system to accurately assess the current heat penetration state. The recalculation of the expected heat penetration depth is based on the updated parameter set, substituting the expected residence time at the current velocity into a linear function to obtain the theoretical heat penetration depth value.

[0071] Understandably, the evaluation of the verification results employs a multi-indicator comprehensive method. Measured values ​​are obtained through thermal imaging and temperature probes. Comparison with expected values ​​considers not only absolute error but also the distribution characteristics of the error. If the error exhibits a systematic shift, further correction is required; if the error is randomly distributed, it is considered basically accurate. Verification results include statistical measures such as the mean error, standard deviation, and maximum deviation. These data are recorded and used to evaluate the effectiveness of adjustments. Once verification is successful, the new correlation is formally adopted, replacing the original correspondence, completing one optimization iteration.

[0072] Step S106: Extract the heat penetration trend from the adjusted correlation. When the heat penetration trend matches the standard curing requirements, obtain the temperature setpoints for all temperature zones inside the tunnel oven to confirm that the internal temperature distribution of the product has reached a uniform level.

[0073] The slope change sequence and intercept adjustment record are extracted from the adjusted correlation. A refined continuous curve showing the heat penetration depth over time is obtained by linear interpolating the sequence at fixed time intervals. The heat penetration trend characteristics are identified based on the positive or negative change of the curve slope. The heat penetration trend characteristics are compared point-by-point with the ripening temperature standard corresponding to the product type. If the value of the trend curve at the end of baking reaches the preset ripening depth threshold, a successful match is determined. The temperature setpoints for all temperature zones within the tunnel oven are obtained at this time, and the real-time temperature readings for each zone are recorded. The temperature values ​​of the product's surface and middle layers are calculated using heat conduction based on the real-time temperature readings. The coefficient of variation is obtained by dividing the standard deviation of each layer's temperature value by the average value. If the coefficient of variation is lower than the preset uniformity threshold, the internal temperature distribution of the product is confirmed to be uniform.

[0074] Specifically, the slope change values ​​over a continuous time period are extracted from the adjusted data on the relationship between heat penetration depth and residence time. These slopes reflect the dynamic changes in heat transfer efficiency under different production conditions.

[0075] Specifically, the new slope value generated after each parameter adjustment forms a change sequence with the previous slope value, recorded every 30 seconds. The intercept adjustment record reflects the changing trend of the product's initial heat penetration state. When the product's moisture content or density changes, the intercept parameter will be adjusted accordingly to adapt to the new initial heat conduction conditions. Linear interpolation processing uses a fixed 5-second time interval to process the discrete sequence data into a continuous process. The interpolation algorithm establishes a linear relationship between two adjacent data points, calculates the heat penetration depth value at intermediate moments, and forms a smooth and continuous change curve.

[0076] For example, if the heat penetration depth is 2.5 mm at the 10th second and 4.2 mm at the 20th second, the depth at the 15th second can be calculated as 3.35 mm through linear interpolation. This refinement process can accurately capture the instantaneous changes in the heat penetration process.

[0077] It should be noted that the heat penetration trend characteristic identification is achieved through the positive and negative changes in the curve slope. The first derivative of the continuous curve at each time point is calculated; a positive slope indicates that the heat penetration depth is increasing, and heat is continuously transferred into the product; a negative slope indicates that the heat penetration process is stabilizing or beginning to slow down. The entire baking process is divided into three typical stages using feature identification: a rapid penetration phase with a slope greater than 0.3 mm / s, a stable penetration phase with a slope between 0.1 and 0.3 mm / s, and a slowing penetration phase with a slope less than 0.1 mm / s. By identifying the transition points and durations of these stages, the current state of the heat treatment process can be accurately determined.

[0078] In one embodiment, the ripening temperature standard corresponding to the product type is formulated based on the process requirements of different baked goods. For bread products, the ripening standard requires an internal temperature of 85°C or higher for at least 3 minutes, with a heat penetration depth reaching 90% of the product thickness. For biscuit products, due to their lower moisture content, the ripening temperature standard is an internal temperature of 75°C and a heat penetration depth of 95% of the thickness. The standard for cake products is relatively lenient; an internal temperature of 70°C and a heat penetration depth of 85% are sufficient. The real-time monitored heat penetration trend curve is compared point-by-point with these preset standards, and the difference between the value at the end of the curve and the target threshold is used to determine whether the ripening requirements have been met. A multi-verification mechanism is used to ensure accuracy in determining successful matching. In addition to the heat penetration depth reaching the preset threshold, the stability characteristics of the curve are also checked, requiring that the change in heat penetration depth not exceed 0.5 mm within the last 30 seconds of baking, indicating that the heat conduction process is basically complete. When a successful match is determined, the current set values ​​of all temperature zones inside the tunnel furnace are immediately obtained, including parameter configurations such as 180°C for the preheating zone, 220°C for the heating zone, 200°C for the heat preservation zone, and 150°C for the cooling zone. At the same time, the real-time temperature readings of each zone are recorded to establish a parameter file for the successful process.

[0079] Specifically, the calculation of the multi-layer temperature distribution is based on a combination of one-dimensional heat conduction theory and actual measurement data. The surface temperature is directly measured by an infrared temperature sensor installed at the top of the tunnel furnace, with an accuracy of ±2℃. The middle layer temperature is located at 50% of the product thickness and is calculated using the heat conduction equation Tmiddle = Tsurface × exp(-αt / δ). 2 The calculation yields the result, where α is the thermal diffusivity, t is the heating time, and δ is the preset thickness coefficient. This layered calculation method accurately reflects the heat propagation pattern and temperature gradient distribution within the product. The coefficient of variation (CV) is calculated using statistical methods to quantify the uniformity of temperature distribution. First, the arithmetic mean of the temperature values ​​at the surface and middle layers is calculated. Then, the standard deviation is calculated to reflect the dispersion of the temperature distribution. The CV value is obtained by dividing the standard deviation by the mean. When the CV value is below 0.15, it indicates that the internal temperature distribution of the product is relatively uniform, and the heat conduction process is well controlled. When the CV value exceeds 0.25, it indicates the existence of a significant temperature gradient, which may lead to uneven baking and quality problems.

[0080] Understandably, the uniformity threshold is set to account for the differences in product characteristics and quality requirements. For high-quality refined bread products, the uniformity threshold is set at 0.12 to ensure consistent product texture; for ordinary baked goods, the threshold can be relaxed to 0.18 to improve production efficiency while ensuring basic quality. When the actual calculated coefficient of variation is lower than the corresponding threshold, it is confirmed that the internal temperature distribution of the product has reached the expected level of uniformity. At this point, the process parameter configuration is recorded and prioritized for the production control of subsequent similar products.

[0081] Step S107: Based on the cumulative data of the confirmed uniformity level and running speed change records, determine the duration data of the new baked product in each temperature zone.

[0082] Based on the confirmed uniformity level values ​​and accumulated data from the recorded speed variations, energy consumption records under the same uniformity level are queried from the historical database. The energy consumption values ​​corresponding to different speed variation patterns are compared to identify the speed and temperature parameter combinations with the lowest energy consumption, thus determining the direction for continuous improvement in energy consumption optimization and control. Using the parameter configuration corresponding to this improvement direction, the current position information of the product and the real-time temperature distribution status of each area are obtained. By spatially mapping the product coordinates to the temperature sensor positions, a temperature value mapping for the product's location is established. The heat accumulation trajectory is calculated based on the temperature sequence along the product's movement path. Based on the heat accumulation trajectory and the real-time speed data of the conveyor belt, the time difference between the product's entry and exit from the boundaries of each temperature zone is calculated by recording the time the product enters and leaves the boundaries of adjacent boundaries, thus obtaining new data on the duration of the baked product's stay in each temperature zone.

[0083] Specifically, in one implementation, the establishment and query mechanism of a historical database is the foundation for achieving energy consumption optimization. The database is categorized and stored according to multiple dimensions such as product type, uniformity level, and speed variation pattern. After each production batch, key parameters such as uniformity level values, conveyor belt speed variation curves, temperature setpoints for each zone, and total energy consumption are automatically recorded. During a query, the current uniformity level value is first matched, and the top ten batches with the lowest energy consumption values ​​are selected from historical records of the same uniformity level. The speed and temperature parameter characteristics of these batches are analyzed, and common patterns are extracted as directions for improvement.

[0084] Specifically, speed variation modes are classified into three types: constant speed mode, stepped speed variation mode, and continuous speed variation mode. In constant speed mode, the conveyor belt speed remains constant, and energy consumption mainly depends on the temperature setting. Stepped speed variation mode uses different speeds at different baking stages, requiring coordination between speed switching timing and temperature adjustment. Continuous speed variation mode adjusts the speed in real time according to the product status, demanding the highest control precision. Comparative analysis reveals that stepped speed variation mode typically has lower energy consumption levels while ensuring product quality.

[0085] It should be noted that the spatial correspondence between the product position and the temperature sensor is achieved through coordinate mapping. A three-dimensional coordinate system is established inside the tunnel furnace, and the position coordinates of the temperature sensors are pre-calibrated and stored. The real-time coordinates of the product within the furnace are calculated based on the conveyor belt speed and entry time. The nearest neighbor interpolation method is used to map the product coordinates to the nearest temperature sensor to obtain the temperature value at that location. When the product is located between two sensors, the temperature value is calculated using a distance-weighted average.

[0086] For example, the calculation of the heat accumulation trajectory involves a time integration process. From the moment the product enters the tunnel oven, the system records its position and corresponding temperature value at each sampling cycle, calculates the heat absorbed during that time period, and accumulates it. The amount of heat absorbed is equal to the product of temperature and time interval multiplied by the thermal conductivity coefficient. As the product moves within the oven, a cumulative curve of heat change with position is formed, i.e., the heat accumulation trajectory. This trajectory reflects the product's heating history throughout the baking process.

[0087] Preferably, the boundaries of temperature zones are determined based on temperature gradient changes. Significant temperature differences exist between adjacent zones, and the zone boundaries are determined by detecting abrupt changes in the temperature gradient. The times when products enter and leave each zone are determined by comparing their position coordinates with the boundary coordinates. The time recording accuracy reaches 0.1 seconds, ensuring the accuracy of duration calculations. The new duration data is compared with the original set values ​​to provide optimization references for the next batch of production.

[0088] Step S108: Based on the new dwell time data, identify the stability level of the current operating speed of the conveyor belt and obtain a comprehensive heat penetration depth equilibrium result as the end mark of the control.

[0089] The actual dwell time values ​​for each temperature zone are extracted from the new dwell time data. The dwell time deviation rate is obtained by calculating the ratio of the actual dwell time to the initial set dwell time. The heat conduction optimization index is obtained by multiplying the deviation rate by the temperature difference of the zone. Using the heat conduction optimization index as a weighting coefficient, the speed sequence of the conveyor belt within a preset time period is obtained. By calculating the variance and coefficient of variation of the speed sequence, if the coefficient of variation is lower than a preset stability threshold, the current operating speed of the conveyor belt is determined to have reached a stable level. Based on the stability level and the heat conduction optimization index, the deviation between the heat penetration value at different depth positions of the product and the standard curing depth is calculated. By accumulating the absolute values ​​of the deviations at all depth positions and dividing by the total number of depth sampling points, a comprehensive heat penetration depth equilibrium result is obtained as the end mark of the control.

[0090] Specifically, in one implementation, the duration deviation rate is calculated based on a comparison between actual operating data and the initial settings. The initial set duration is predetermined before production begins based on the product formula and process requirements. Typically, bread products spend 60-90 seconds in the preheating zone, 180-240 seconds in the main baking zone, and 60-120 seconds in the post-baking zone. The actual dwell time is recorded in real time by a product position tracking system with an accuracy of 0.1 seconds. The deviation rate is equal to the difference between the actual duration and the set duration divided by the set duration; a positive value indicates an extended dwell time, and a negative value indicates a shortened dwell time.

[0091] Specifically, the calculation of the heat conduction optimization index comprehensively considers both time and temperature dimensions. This index, acting as a weighting coefficient, adjusts the importance of conveyor belt speed monitoring; a larger index value indicates a more severe deviation in heat conduction, requiring closer monitoring of speed stability, which can be achieved by increasing the sampling frequency and analysis precision of speed data. The regional temperature difference is obtained by subtracting the average temperature of adjacent regions, reflecting the magnitude of the temperature gradient. The product of the time deviation rate and the temperature difference characterizes the degree to which the actual heat transfer effect deviates from expectations. When the deviation rate is positive and the temperature difference is large, it indicates that the product has remained in the high-temperature gradient region for too long, potentially leading to surface overheating; conversely, it may result in insufficient heating. This index provides a weighting basis for subsequent stability assessments.

[0092] It should be noted that the conveyor belt speed sequence acquisition covers the entire baking cycle. The preset time period is typically 5 minutes, containing 300 speed sampling points. Variance calculation reflects the absolute amplitude of speed fluctuations; the variance calculation formula is as follows: Where vi is the i-th velocity value. The average speed is given by , and n is the total number of sampling points. Variance reflects the absolute amplitude of speed fluctuations; a larger value indicates more drastic speed changes, which may lead to uneven heating of the product. Coefficient of Variation. The standard deviation divided by the average speed eliminates the influence of the absolute speed value, allowing for an objective assessment of relative stability at different speed levels. The stability threshold is set according to the product type; thicker products are allowed a higher coefficient of variation, while thinner products require more stringent speed stability. A stable speed is considered to have been reached when the coefficient of variation is below 0.05.

[0093] For example, a multi-level sampling method is used to measure the heat penetration depth. Five to seven depth sampling points are set at equal intervals from the surface to the center of the product. The temperature value of each point is obtained through thermocouples or infrared thermometry to estimate the degree of heat penetration at that depth. The standard curing depth is the penetration depth that should be achieved when the product is fully cured, and it is predetermined based on the product thickness and density. The calculation of the depth position deviation first converts the stability level into a stability coefficient K = 1 - CV, and then multiplies it by the heat conduction optimization index H to obtain a comprehensive adjustment factor F = K × H. The measured heat penetration value at each depth position is multiplied by the adjustment factor F and compared with the standard curing depth; the deviation value = |measured value × F - standard value|. The absolute value of the deviation at each sampling point reflects the degree of local non-uniformity; the average deviation is obtained by summing the deviations and dividing by the total number of sampling points.

[0094] Preferably, the determination of the end of regulation adopts a dual standard. In addition to the requirement that the heat penetration depth equalization result be lower than a preset threshold, it is also required that the equalization result remain stable for three consecutive sampling cycles, with a change of no more than 5%. This dual determination mechanism avoids misjudgment caused by instantaneous fluctuations, ensuring that the system only ends regulation when it truly reaches a stable equalization state, thus achieving precise control of the baking process.

[0095] Specific embodiments of the present invention have been described above. It should be understood that the present invention is not limited to the specific embodiments described above, and those skilled in the art can make various changes or modifications within the scope of the claims, which do not affect the essence of the present invention. Unless otherwise specified, the embodiments and features described in this application can be arbitrarily combined with each other.

Claims

1. A method for automatic control of energy consumption optimization in a tunnel furnace, characterized in that, include: Acquire information on the running speed of the tunnel oven conveyor belt and the position of the baked products, extract heat penetration depth data from the temperature distribution, identify changes in running speed, and evaluate the trend deviation value of the impact of changes in running speed on the uniformity of heat penetration. Based on the aforementioned trend deviation value and residence time data, the heat conduction status of the product in each temperature zone is determined, and an initial heat penetration index is obtained. If the initial heat penetration index is lower than the standard value, the temperature setpoint of the conveyor belt feed end temperature zone is adjusted. Adjustment amplitude data is extracted from the adjusted temperature setpoint to assess the heating risk of the conveyor belt discharge end temperature zone, and the temperature setpoint of the discharge end temperature zone is adjusted accordingly. The adjusted temperature setpoint of the discharge end temperature zone is used to assess the total adjustment amount of the temperature setpoints in each zone, generating a temperature configuration combination. Based on the energy consumption control scheme and conveyor belt speed monitoring data, the correlation between heat penetration and residence time is determined. The heat penetration trend is extracted from the correlation. When it matches the standard curing requirements, the temperature setpoints of all temperature zones are obtained, confirming that the internal temperature distribution of the product has reached a uniform level. After confirming that the internal temperature distribution of the product has reached a uniform level, new residence time data is generated. Based on the new residence time data, the stability of the conveyor belt running speed is identified, generating a heat penetration depth equilibrium result.

2. The automatic control method for optimizing energy consumption of a tunnel furnace according to claim 1, characterized in that, The process of acquiring information on the conveyor belt speed and the position of the baked products in the tunnel oven, extracting heat penetration depth data from the temperature distribution, identifying changes in operating speed, and evaluating the trend deviation value of the impact of changes in operating speed on the uniformity of heat penetration includes: The system acquires real-time conveyor belt speed data and the position coordinates of the baked product inside the oven, collects surface temperature distribution data of the product, calculates the heat penetration depth from the product surface to the interior, and records the amplitude and frequency of the conveyor belt speed change within a time window. Based on the amplitude and frequency of the change, the system calculates the dwell time of the product in each temperature zone, extracts temperature values ​​at different depths of the product from the temperature distribution data, calculates the temperature gradient distribution, compares it with a standard baking curve to determine the degree of deviation of the heat penetration depth, and generates an influence coefficient. The system then multiplies the influence coefficient by the heat received by the product to generate a heat penetration adjustment value for each zone, predicts the trend of heat penetration depth change at the next moment, and calculates the trend deviation value.

3. The automatic control method for optimizing energy consumption of a tunnel furnace according to claim 1, characterized in that, The step of determining the heat conduction status of the product in each temperature zone based on the influence trend deviation value and residence time data, obtaining the initial heat penetration index, and adjusting the temperature setpoint of the conveyor belt feed end temperature zone when the initial heat penetration index is lower than the standard value includes: Based on the aforementioned influence trend deviation value and residence time data, the heat received by the product in each temperature zone is calculated, the heat conduction rate is determined, and heat conduction state distribution data is generated. The initial temperature value of the product is extracted from the product location information, the temperature zones traversed by the product and the duration of heating are traced, and the initial heat penetration index is calculated. The initial heat penetration index is compared with the standard heat penetration curve to determine the temperature adjustment range and generate the corrected temperature setpoint for the feed end temperature zone. The feed end heating power configuration is updated according to the corrected temperature setpoint, the temperature increment is allocated, and the adjusted feed end temperature zone temperature setpoint is generated.

4. The automatic control method for optimizing energy consumption of a tunnel furnace according to claim 1, characterized in that, The step of extracting adjustment range data from the adjusted temperature setpoint, assessing the heating risk in the temperature zone at the discharge end of the conveyor belt, and adjusting the temperature setpoint in the temperature zone at the discharge end includes: The difference between the current temperature value and the original temperature value is extracted from the adjusted temperature setpoint of the feed end temperature zone to generate adjustment amplitude data, calculate the additional heat accumulation value, compare it with the safe heat threshold, and determine the heating risk level; the temperature reduction coefficient is determined according to the risk level, the temperature reduction amount at the discharge end is calculated, and it is allocated to the discharge end heating unit; the discharge end heating control parameters are updated according to the temperature reduction amount, the heat supply balance is verified, and the adjusted temperature setpoint of the discharge end temperature zone is generated.

5. The automatic control method for optimizing energy consumption of a tunnel furnace according to claim 1, characterized in that, The process of obtaining the adjusted temperature setpoint of the discharge end temperature zone, evaluating the total adjustment of the temperature setpoint of each zone, and generating a temperature configuration combination includes: The adjusted temperature setpoint of the discharge end temperature zone and the real-time readings of each temperature sensor are obtained. The total adjustment amount is obtained by accumulating the absolute values ​​of the adjustment amplitudes of all zones. The power consumption value corresponding to each temperature is queried from the energy consumption records established from historical production data. The temperature setpoint of each zone is iteratively optimized using a genetic algorithm with the goal of minimizing total energy consumption and the constraint of ensuring that the product center temperature reaches the preset ripening temperature. The temperature configuration combination with the minimum energy consumption is searched.

6. The automatic control method for optimizing energy consumption of a tunnel furnace according to claim 1, characterized in that, The step of determining the correlation between heat penetration and residence time based on the temperature configuration combination and conveyor belt speed monitoring data includes: Based on the temperature configuration combination and conveyor belt speed monitoring data, the real-time position coordinates of the product are obtained, the current heat penetration depth value is calculated, a depth value sequence is generated, and the changing trend is identified; the changing trend is compared with the ideal heat penetration trajectory, the deviation value is calculated, and the correlation between heat penetration depth and residence time is fitted.

7. The automatic control method for optimizing energy consumption of a tunnel furnace according to claim 1, characterized in that, The step of extracting the heat penetration trend from the correlation and matching it with the standard curing requirements, obtaining the temperature setpoints for all temperature zones, and determining that the internal temperature distribution of the product has reached a uniform level includes: Extract the slope change sequence from the correlation to generate a continuous curve of heat penetration depth changing with time and identify trend characteristics; compare the trend characteristics with the aging temperature standard to determine the matching status, obtain the temperature setpoint and real-time temperature readings of all temperature areas; calculate the temperature values ​​of each layer of the product based on the real-time temperature readings to confirm that the internal temperature distribution of the product has reached a uniform level.

8. The automatic control method for optimizing energy consumption of a tunnel furnace according to claim 1, characterized in that, After determining that the internal temperature distribution of the product has reached a uniform level, new residence time data is generated, including: Based on the uniformity level and operating speed change data, query historical energy consumption records, identify the parameter combination with the lowest energy consumption, and determine the direction of optimization control; based on the optimization control direction, obtain product location information and temperature distribution status, establish temperature value mapping, and calculate heat accumulation trajectory; based on the heat accumulation trajectory and conveyor belt speed data, calculate the dwell time of the product in each temperature zone, and generate new dwell time data.

9. The automatic control method for optimizing energy consumption of a tunnel furnace according to claim 1, characterized in that, The step of identifying the conveyor belt speed stability based on the new dwell time data and generating a heat penetration depth equalization result includes: The dwell time values ​​of each region are extracted from the new dwell time data, the conveyor belt speed sequence is obtained, the coefficient of variation is calculated, the stability of the running speed is judged, the thermal penetration value deviation is calculated, and the thermal penetration depth equilibrium result is generated.

Images