A multi-sensor fusion-based precise temperature control method and system for a commercial oven

Abstract

This application relates to the field of oven temperature control technology, and in particular to a method and system for precise temperature control of commercial ovens based on multi-sensor fusion. The method includes: acquiring multi-source sensor data from the commercial oven; temperature field modeling: dividing the oven cavity into multiple temperature field regions, establishing a temperature field state model characterizing the temperature of each temperature field region, and establishing a measurement mapping relationship between the multi-source sensor data and the temperature field state model; obtaining the temperature field estimation result of the oven cavity based on the temperature field state model and the measurement mapping relationship; identifying door opening events based on door state data, and estimating the door opening disturbance based on the temperature field estimation result; switching the temperature control process to an opening stage and a closing recovery stage based on the door opening event, obtaining the temperature control quantity; and ensuring the oven cavity temperature field meets the set temperature control target based on the temperature control quantity. This application effectively prevents large-scale heat loss and effectively suppresses temperature overshoot, ensuring the consistency of the temperature field throughout the baking process.

Description

Technical Field

[0001] This application relates to the field of oven temperature control technology, and in particular to a method and system for precise temperature control of commercial ovens based on multi-sensor fusion. Background Technology

[0002] Commercial ovens (hot air ovens, rack ovens, combination ovens, etc.) have the following common problems in actual operation: Traditional temperature control relies on a single point or a small number of measuring points (such as a thermocouple at a certain point in the cavity), which cannot reflect the true temperature distribution inside the oven cavity, resulting in uneven baking, under-baking or over-baking at the edges and corners. Opening the oven door to load / check is a frequent operation. Opening the door will cause significant heat loss and cold air rush in, resulting in a sudden drop in temperature, slow recovery, or recovery overshoot. Existing controls mostly use simple PID or fixed compensation, lacking quantitative prediction and adaptive compensation for disturbances such as door opening duration, door opening angle, and load conditions, making it difficult to balance temperature control accuracy and consistency. Summary of the Invention

[0003] This application provides a method and system for precise temperature control of commercial ovens based on multi-sensor fusion to solve the above-mentioned problems.

[0004] In a first aspect, this application provides a method for precise temperature control of a commercial oven based on multi-sensor fusion, the method comprising: Data acquisition: Acquire multi-source sensor data of the commercial oven, including multi-point temperature data of the oven cavity, actuator status data, and door status data; Temperature field modeling: The oven cavity of the commercial oven is divided into multiple temperature field regions, a temperature field state model characterizing the temperature of each temperature field region is established, and a measurement mapping relationship between the multi-source sensor data and the temperature field state model is established. Temperature field inversion: Based on the temperature field state model and the measurement mapping relationship, the multi-source sensor data is fused and estimated to obtain the temperature field estimation result of the furnace cavity; Door opening disturbance analysis: Identify door opening events based on the door status data, and estimate the door opening disturbance amount based on the temperature field estimation results; Phased compensation control: Based on the door opening event, the temperature control process is switched to a door opening stage and a door closing recovery stage. In the door opening stage, the heating output is suppressed and controlled according to the door opening disturbance. In the door closing recovery stage, a feedforward compensation control quantity is generated based on the door opening disturbance and superimposed with the feedback control quantity generated based on the temperature field estimation result to obtain the temperature control quantity. Execution control: Control the temperature control actuator of the commercial oven according to the temperature control quantity, so that the oven cavity temperature field meets the set temperature control target.

[0005] Through the above technical solutions, the system can monitor the real-time temperature distribution inside the oven cavity, avoiding the localized temperature difference problems caused by traditional single-point temperature measurement. Door opening disturbance analysis accurately quantifies heat loss and cold air intrusion caused by door opening, enabling the phased compensation control module to implement refined control strategies. Specifically, the suppression control during the door opening phase effectively prevents large-scale heat loss, while the feedforward control command generation during the door closing recovery phase provides rapid energy replenishment based on the disturbance, significantly shortening the temperature recovery time. Simultaneously, by superimposing with the feedback control command generation, it effectively suppresses temperature overshoot, ensuring temperature consistency throughout the entire baking process.

[0006] Optionally, the temperature field modeling includes: Based on the physical structure of the oven cavity of a commercial oven, the internal space of the oven cavity is geometrically discretized into N interrelated temperature field regions, where N is an integer greater than 2; A temperature field state vector is defined to characterize the overall temperature distribution of the furnace cavity at the current moment. The temperature field state vector is composed of N equivalent temperature values ​​of the temperature field regions arranged sequentially. By using historical test data of the oven, a system matrix describing the thermal coupling relationship between each temperature field region and a control matrix describing the influence of the temperature control actuator on the temperature of each temperature field region were calibrated and obtained. Based on the physical installation positions of all temperature sensors in the furnace cavity, the mapping weight between the measured value of each temperature sensor and the temperature of one or more of the temperature field regions is determined, forming a measurement mapping matrix, and completing the establishment of the mapping relationship from the temperature field state vector to the actual collectable multi-point temperature data.

[0007] Through the aforementioned technical solution and precise temperature field modeling, the system can capture the non-uniformity of temperature distribution within the furnace cavity, rather than relying solely on single-point measurements. The establishment of the measurement mapping matrix ensures that the measurements from a limited number of temperature sensors can reflect the changes in the entire temperature field state vector to the greatest extent possible. This provides high-precision prior information for data fusion in the subsequent temperature field inversion module, significantly improving the accuracy of estimating the true temperature field state and solving the problem that measurement points cannot represent the global picture in traditional solutions.

[0008] Optionally, the temperature field inversion includes: Based on the temperature field state model, initialize the estimated value of the temperature field state vector and its estimation error covariance; The multi-point temperature data acquired within the same sampling period are time-aligned and filtered and denoised to obtain the observation vector at the current moment. Using the temperature field estimation results from the previous moment, the actuator state data at the current moment, and the system matrix and control matrix in the temperature field state model, the prior estimate of the temperature field state at the current moment is predicted. The observation vector at the current moment is compared with the predicted observation value calculated by the prior estimate through the measurement mapping matrix. The observation residual is calculated, and the prior estimate is corrected based on the Kalman gain to obtain the optimal temperature field estimation result at the current moment, that is, the fused estimate of the temperature of each temperature field region in the furnace cavity.

[0009] By employing the aforementioned technical solution and using Kalman filtering for temperature field inversion, discrete and noisy sensor measurements can be effectively fused into a continuous, physical model-based temperature field state. This dynamic fusion method not only provides temperature estimates for the temperature field grid outside the sensor mounting points but also tracks rapid changes in the internal temperature field of the furnace cavity in real time. Particularly for nonlinear thermal disturbances that traditional control methods cannot detect, it provides high-precision, low-latency temperature field state information, laying a solid foundation for accurate decision-making in subsequent staged compensation control modules.

[0010] Optionally, the door opening disturbance analysis includes: The door status data is monitored in real time. When the door changes from closed to open or the door opening degree exceeds a preset threshold, an opening event is determined to have occurred, and the opening start time is recorded. During the duration of the door opening event, based on the temperature field estimation results, the rate of decrease of the overall average temperature of the furnace cavity and / or the trend of the maximum temperature difference between each temperature field region are calculated in real time. Based on the duration of the door opening event, the door opening characteristics, the rate of decrease of the overall average temperature of the furnace cavity, and the ambient temperature, the heat loss caused by the door opening is quantitatively calculated using a pre-established heat loss lookup table. At the same time, based on the trend of the maximum temperature difference change, the intensity of cold air intrusion is assessed, which together constitute the door opening disturbance.

[0011] Through the above technical solution and refined disturbance quantification, this method solves the problem of inaccurate estimation of heat loss during door opening in traditional solutions. The door opening disturbance not only includes the total energy loss but also distinguishes the temperature field non-uniformity caused by cold air intrusion, enabling subsequent compensation control to provide targeted energy replenishment and temperature field homogenization.

[0012] Optionally, the phased compensation control includes: The start time of the door opening event is taken as the trigger point for entering the door opening stage, and the moment when the door returns to the closed state is taken as the trigger point for entering the door closing and recovery stage. During the door opening phase, the output power of the heating actuator is limited to a level below the rated power, and the fan speed or damper opening is dynamically adjusted according to the real-time estimate of the door opening disturbance to reduce heat loss due to forced convection. During the door closing recovery phase, the energy compensation requirement calculated based on the door opening disturbance is converted into feedforward control commands for one or more temperature control actuators. At the same time, based on the deviation between the temperature field estimation result and the set temperature control target, feedback control commands are generated through a feedback control algorithm. The feedforward control command and the feedback control command are superimposed and then subjected to physical limiting constraints by the actuator to generate the final temperature control quantity.

[0013] The advantage of the phased compensation control strategy, achieved through the above technical solution, lies in its ability to "predict" and "respond quickly" to disturbances. The suppression control during the door opening phase effectively reduces heat loss, while the superimposed control of feedforward and feedback during the door closing recovery phase allows the system to inject compensation energy in advance before the temperature drops significantly, greatly shortening the temperature recovery time. At the same time, the feedback control corrects the residual errors introduced by the model and sensors, ensuring the smoothness of the recovery process and the final temperature control accuracy.

[0014] Optionally, during the process of geometric discretization of the internal space of the furnace cavity, the specific method for dividing the temperature field region is as follows: the furnace cavity is divided into three layers (upper, middle, and lower) along the height direction, into two rows (front and rear) along the depth direction, and into two sides (left and right) along the width direction. The furnace cavity is divided into at least six independent temperature field regions through the combination of three-dimensional space.

[0015] By employing the aforementioned technical solution and utilizing a precise discretization method, the temperature field inversion module can output temperature field estimation results with high spatial resolution, significantly outperforming coarse models that only divide the area into two or three regions. By capturing the temperature states of high-heat-loss regions such as the "lower front layer," it provides necessary data support for the subsequent phased compensation control module to implement non-uniform, regional compensation, thereby improving the uniformity of the temperature field.

[0016] Optionally, the heat loss lookup table is constructed based on the following: the heat loss is positively correlated with the door opening duration, positively correlated with the temperature difference inside and outside the furnace cavity, and positively correlated with the door opening degree; the cold air intrusion intensity is positively correlated with the rate of change of the pressure difference inside and outside the furnace cavity and the trend of the maximum temperature difference.

[0017] Through the above technical solution, the lookup table construction based on physical mechanisms enables the door opening disturbance analysis module to provide high-precision quantification results, avoiding errors caused by simply relying on empirical formulas. Accurate disturbance quantification is the foundation for generating feedforward control commands in the staged compensation control module, thus ensuring that energy compensation during the door closing recovery phase is neither excessive nor insufficient.

[0018] Optionally, the step of dynamically adjusting the fan speed or damper opening based on the real-time estimate of the door opening disturbance includes: Based on the real-time estimated cold air intrusion intensity, when the intrusion intensity exceeds the limit, the fan speed is reduced and the damper is adjusted to reduce the opening of the furnace cavity and the outside, so as to slow down the cold air exchange rate. When the intrusion intensity does not exceed the limit, the fan speed is maintained to keep the airflow in the furnace cavity stable and avoid aggravating temperature stratification.

[0019] The above technical solution further refines the specific strategy for implementing suppression control during the door opening phase using the phased compensation control module. This involves dynamically adjusting the fan and damper to optimize the thermodynamic environment during the door opening process. This feature aims to address the issue in forced convection ovens where fan operation may accelerate cold air intrusion and heat loss.

[0020] Optionally, converting the energy compensation requirement calculated based on the door opening disturbance into feedforward control commands for one or more temperature control actuators includes: Based on the quantified heat loss, it is equivalent to the total energy that needs to be replenished; Based on the difference in temperature drop in each temperature field region during the opening stage, the total energy is non-uniformly distributed to different heating execution circuits. This increases the feedforward power ratio of the heaters corresponding to temperature field regions where the temperature drop exceeds the limit, while simultaneously increasing the feedforward command of the fan speed to enhance the stirring and mixing of hot air in the furnace cavity.

[0021] The above technical solution utilizes a non-uniform feedforward compensation strategy to ensure spatially targeted energy replenishment during the door-closing recovery phase. It not only provides rapid energy replenishment (an advantage of feedforward) but also resolves the temperature field non-uniformity problem estimated by Kalman filtering. By allocating more power to the cold spot region, the system can significantly shorten the time for the overall temperature field to reach uniform stability. Compared to uniform compensation or simple feedback control, the recovery speed is faster and the temperature difference is smaller.

[0022] Secondly, this application provides a commercial oven precision temperature control system based on multi-sensor fusion, the system comprising: The data acquisition module is used to acquire multi-source sensor data of the commercial oven, including multi-point temperature data of the oven cavity, actuator status data and door status data. The temperature field modeling module is used to divide the oven cavity of the commercial oven into multiple temperature field regions, establish a temperature field state model characterizing the temperature of each temperature field region, and establish a measurement mapping relationship between the multi-source sensor data and the temperature field state model. The temperature field inversion module is used to perform fusion estimation on the multi-source sensor data based on the temperature field state model and the measurement mapping relationship to obtain the temperature field estimation result of the furnace cavity; The door opening disturbance analysis module is used to identify door opening events based on the door status data and estimate the door opening disturbance amount based on the temperature field estimation results. The phased compensation control module is used to switch the temperature control process into an opening stage and a closing recovery stage based on the door opening event. In the opening stage, the heating output is suppressed and controlled according to the door opening disturbance. In the closing recovery stage, a feedforward compensation control quantity is generated based on the door opening disturbance and superimposed with the feedback control quantity generated based on the temperature field estimation result to obtain the temperature control quantity. The execution control module is used to control the temperature control actuator of the commercial oven according to the temperature control quantity, so that the oven cavity temperature field meets the set temperature control target.

[0023] Optionally, the temperature field modeling module is specifically used for: Based on the physical structure of the oven cavity of a commercial oven, the internal space of the oven cavity is geometrically discretized into N interrelated temperature field regions, where N is an integer greater than 2; A temperature field state vector is defined to characterize the overall temperature distribution of the furnace cavity at the current moment. The temperature field state vector is composed of N equivalent temperature values ​​of the temperature field regions arranged sequentially. By using historical test data of the oven, a system matrix describing the thermal coupling relationship between each temperature field region and a control matrix describing the influence of the temperature control actuator on the temperature of each temperature field region were calibrated and obtained. Based on the physical installation positions of all temperature sensors in the furnace cavity, the mapping weight between the measured value of each temperature sensor and the temperature of one or more of the temperature field regions is determined, forming a measurement mapping matrix, and completing the establishment of the mapping relationship from the temperature field state vector to the actual collectable multi-point temperature data.

[0024] Optionally, the temperature field inversion module is specifically used for: Based on the temperature field state model, initialize the estimated value of the temperature field state vector and its estimation error covariance; The multi-point temperature data acquired within the same sampling period are time-aligned and filtered and denoised to obtain the observation vector at the current moment. Using the temperature field estimation results from the previous moment, the actuator state data at the current moment, and the system matrix and control matrix in the temperature field state model, the prior estimate of the temperature field state at the current moment is predicted. The observation vector at the current moment is compared with the predicted observation value calculated by the prior estimate through the measurement mapping matrix. The observation residual is calculated, and the prior estimate is corrected based on the Kalman gain to obtain the optimal temperature field estimation result at the current moment, that is, the fused estimate of the temperature of each temperature field region in the furnace cavity.

[0025] Optionally, the door opening disturbance analysis module is specifically used for: The door status data is monitored in real time. When the door changes from closed to open or the door opening degree exceeds a preset threshold, an opening event is determined to have occurred, and the opening start time is recorded. During the duration of the door opening event, based on the temperature field estimation results, the rate of decrease of the overall average temperature of the furnace cavity and / or the trend of the maximum temperature difference between each temperature field region are calculated in real time. Based on the duration of the door opening event, the door opening characteristics, the rate of decrease of the overall average temperature of the furnace cavity, and the ambient temperature, the heat loss caused by the door opening is quantitatively calculated using a pre-established heat loss lookup table. At the same time, based on the trend of the maximum temperature difference change, the intensity of cold air intrusion is assessed, which together constitute the door opening disturbance.

[0026] Optionally, the staged compensation control module is specifically used for: The start time of the door opening event is taken as the trigger point for entering the door opening stage, and the moment when the door returns to the closed state is taken as the trigger point for entering the door closing and recovery stage. During the door opening phase, the output power of the heating actuator is limited to a level below the rated power, and the fan speed or damper opening is dynamically adjusted according to the real-time estimate of the door opening disturbance to reduce heat loss due to forced convection. During the door closing recovery phase, the energy compensation requirement calculated based on the door opening disturbance is converted into feedforward control commands for one or more temperature control actuators. At the same time, based on the deviation between the temperature field estimation result and the set temperature control target, feedback control commands are generated through a feedback control algorithm. The feedforward control command and the feedback control command are superimposed and then subjected to physical limiting constraints by the actuator to generate the final temperature control quantity.

[0027] Optionally, in the temperature field modeling module, during the process of geometric discretization of the internal space of the furnace cavity, the specific method for dividing the temperature field region is as follows: along the height direction of the furnace cavity, it is divided into three layers: upper, middle and lower; along the depth direction of the furnace cavity, it is divided into two rows: front and back; and along the width direction of the furnace cavity, it is divided into left and right sides. Through the combination of three-dimensional space, the furnace cavity is divided into at least six independent temperature field regions.

[0028] Optionally, in the door opening disturbance analysis module, the heat loss lookup table is constructed based on the following: the heat loss is positively correlated with the door opening duration, positively correlated with the temperature difference inside and outside the furnace cavity, and positively correlated with the door opening degree; the cold air intrusion intensity is positively correlated with the rate of change of the pressure difference inside and outside the furnace cavity and the trend of the maximum temperature difference.

[0029] Optionally, when the phased compensation control module dynamically adjusts the fan speed or damper opening based on the real-time estimated value of the door opening disturbance, it is specifically used for: Based on the real-time estimated cold air intrusion intensity, when the intrusion intensity exceeds the limit, the fan speed is reduced and the damper is adjusted to reduce the opening of the furnace cavity and the outside, so as to slow down the cold air exchange rate. When the intrusion intensity does not exceed the limit, the fan speed is maintained to keep the airflow in the furnace cavity stable and avoid aggravating temperature stratification.

[0030] Optionally, when the phased compensation control module converts the energy compensation requirement calculated based on the door opening disturbance into feedforward control commands for one or more temperature control actuators, it is specifically used for: Based on the quantified heat loss, it is equivalent to the total energy that needs to be replenished; Based on the difference in temperature drop in each temperature field region during the opening stage, the total energy is non-uniformly distributed to different heating execution circuits. This increases the feedforward power ratio of the heaters corresponding to temperature field regions where the temperature drop exceeds the limit, while simultaneously increasing the feedforward command of the fan speed to enhance the stirring and mixing of hot air in the furnace cavity. Attached Figure Description

[0031] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0032] Figure 1 This is a schematic diagram illustrating an application scenario provided in one embodiment of this application; Figure 2A flowchart illustrating a method for precise temperature control of a commercial oven based on multi-sensor fusion, provided as an embodiment of this application; Figure 3 This is a schematic diagram of a commercial oven precision temperature control system based on multi-sensor fusion, provided as an embodiment of this application. Detailed Implementation

[0033] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0034] Furthermore, the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. Additionally, the character " / " in this article, unless otherwise specified, generally indicates that the preceding and following related objects have an "or" relationship.

[0035] The embodiments of this application will now be described in further detail with reference to the accompanying drawings.

[0036] Traditional temperature control strategies for commercial ovens often rely on single or a few measuring points (such as a thermocouple at a certain point in the cavity), which cannot reflect the true temperature distribution inside the oven cavity. This leads to uneven baking, under-baking or over-baking of corners and edges. Furthermore, opening the oven door to load / check food is a frequent operation, and opening the door causes significant heat loss and cold air influx, resulting in a sudden drop in temperature, slow recovery, or over-over-recovery. Based on this, this application provides a method and system for precise temperature control of commercial ovens based on multi-sensor fusion. This method allows the system to monitor the real-time temperature distribution inside the oven cavity, avoiding the localized temperature difference problems associated with traditional single-point temperature measurement. Door opening disturbance analysis precisely quantifies heat loss and cold air intrusion caused by door opening, enabling the staged compensation control module to implement refined control strategies. Specifically, the suppression control during the door opening stage effectively prevents large-scale heat loss, while the feedforward control command generation during the door closing recovery stage provides rapid energy replenishment based on the disturbance, significantly shortening the temperature recovery time. Simultaneously, by superimposing with the feedback control command generation, it effectively suppresses temperature overshoot, ensuring temperature consistency throughout the baking process.

[0037] Figure 1This application provides an illustration of an application scenario. In the operation of a commercial oven, the method provided in this application effectively prevents large-scale heat loss and suppresses temperature overshoot.

[0038] Specifically, the method provided in this application can be applied to any server. The server interacts with a multi-source sensor group and a commercial oven, and obtains multi-source sensor data provided by the multi-source sensor group through the server. Based on the multi-source sensor data, temperature field modeling and door opening disturbance analysis are performed. The heat loss and cold air intrusion caused by door opening are accurately quantified, enabling the staged compensation control module to implement a refined control strategy. Precise temperature control is achieved by providing precise temperature control quantities to the commercial oven. The suppression control during the door opening stage effectively prevents large-scale heat loss, while the feedforward control command generation during the door closing recovery stage provides rapid energy replenishment based on the disturbance quantity, significantly shortening the temperature field recovery time. At the same time, by superimposing with the feedback control command generation, the temperature overshoot phenomenon is effectively suppressed, ensuring the temperature field consistency of the product throughout the baking process.

[0039] For specific implementation details, please refer to the following examples.

[0040] Figure 2 This is a flowchart illustrating a method for precise temperature control of a commercial oven based on multi-sensor fusion, as provided in one embodiment of this application. The method of this embodiment can be applied to servers in the above-mentioned scenarios. Figure 2 As shown, the method includes: S201. Data Acquisition: Acquire multi-source sensor data of the commercial oven, including multi-point temperature data of the oven cavity, actuator status data, and door status data. S202, Temperature field modeling: Divide the oven cavity of the commercial oven into multiple temperature field regions, establish a temperature field state model that characterizes the temperature of each temperature field region, and establish a measurement mapping relationship between multi-source sensor data and temperature field state model. S203. Temperature field inversion: Based on the temperature field state model and measurement mapping relationship, multi-source sensor data are fused and estimated to obtain the temperature field estimation result of the furnace cavity. S204. Door opening disturbance analysis: Identify door opening events based on door status data and estimate the door opening disturbance based on temperature field estimation results; S205, phased compensation control: Based on the door opening event, the temperature control process is switched to the door opening stage and the door closing recovery stage. In the door opening stage, the heating output is suppressed and controlled according to the door opening disturbance. In the door closing recovery stage, the feedforward compensation control quantity is generated based on the door opening disturbance and superimposed with the feedback control quantity generated based on the temperature field estimation result to obtain the temperature control quantity. S206, Execution Control: Control the temperature control actuator of the commercial oven according to the temperature control quantity to ensure that the oven cavity temperature field meets the set temperature control target.

[0041] When operating a commercial oven with the door open, the traditional single-point temperature control system often fails to accurately sense the true temperature distribution and heat loss within the oven cavity. This results in significant overshoot or slow recovery of the temperature after the door is closed, affecting the consistency of baked products.

[0042] The commercial oven used in this embodiment is a circulating hot air oven. The multi-source sensor group includes PT100 platinum resistance temperature sensors installed at different locations inside the oven cavity to collect multi-point temperature data. The temperature control actuator uses a heating element controlled by a PID solid-state relay (SSR) and a variable frequency fan; its status data is collected through an electrical interface. Door status data is obtained from mechanical limit switches. The temperature field state model module discretizes the oven cavity into, for example, four temperature field region grids, constructing a temperature field model based on a state-space model. The temperature field inversion module uses an unscented Kalman filter (UKF) to achieve fusion estimation of multi-source data. When the door opening disturbance analysis module detects that the door is open, it calculates the average temperature drop rate of the oven cavity based on the temperature field estimation results and quantifies the door opening disturbance through a preset heat loss lookup table. The staged compensation control module is responsible for state switching and generating temperature control quantities. During the door opening stage, the temperature control quantity is set to the suppression value of the maintenance power. During the door closing recovery phase, the door opening disturbance input is converted into the energy required to generate the feedforward control command. After being superimposed with the generated feedback control command, it is output to the temperature control actuator to implement precise power regulation.

[0043] In other alternative implementations, the multi-source sensor array can employ K-type thermocouples or NTC thermistors as temperature sensor groups, and even supplement with thermal imagers for surface temperature monitoring in specific areas. Door status data can be obtained using Hall effect sensor signals or encoders to detect door opening characteristics, in addition to mechanical limit switches. The fusion algorithm of the temperature field inversion module can be replaced with an extended Kalman filter (EKF) or an adaptive Kalman filter. The temperature control actuator can be a three-phase heating contactor or a damper adjustment servo motor. The temperature field state model module can discretize the furnace cavity into six, eight, or even more temperature field region grids to improve the accuracy of spatial temperature estimation.

[0044] In some embodiments, temperature field modeling includes: discretizing the internal space of a commercial oven cavity into N interconnected temperature field regions based on the physical structure of the cavity, where N is an integer greater than 2; defining a temperature field state vector to characterize the overall temperature distribution of the cavity at the current moment, which is composed of the equivalent temperature values ​​of the N temperature field regions arranged sequentially; calibrating and obtaining a system matrix describing the thermal coupling relationship between each temperature field region and a control matrix describing the influence of the temperature control actuator on the temperature of each temperature field region using historical test data; and determining the mapping weight between the measured value of each temperature sensor and the temperature of one or more temperature field regions based on the physical installation position of all temperature sensors in the cavity, forming a measurement mapping matrix, and completing the establishment of the mapping relationship from the temperature field state vector to the actual collectable multi-point temperature data.

[0045] This implementation focuses on addressing the problems of model mismatch and uneven temperature field distribution in traditional temperature control. By discretizing the internal space of the furnace cavity into multiple temperature field region grids and constructing a temperature field state model based on a state-space model, the system can transform the physical heat conduction process into a computable linear or nonlinear dynamic system. This conceptual model based on physical decoupling provides the foundation for accurate estimation in the subsequent temperature field inversion module. Specifically, after the system starts up, the geometric discretization of the furnace cavity is first completed through offline testing and thermodynamic analysis, and the system matrix and control matrix are calibrated. These matrices describe the heat transfer between temperature field regions and the influence of actuator power on the temperature of each region when there is no external disturbance or control input.

[0046] In its implementation, the temperature field modeling module discretizes the furnace cavity into N temperature field region grids, for example, N=4 regions, corresponding to the upper, lower, front, and rear parts of the furnace cavity, respectively. The temperature field state vector $X(k)$ is composed of the equivalent temperatures T_1, T_2, T_3, and T_4 of these four regions. Using historical operating data and a system identification method, the system matrix A and control matrix B are calibrated. Simultaneously, based on the physical installation points of multiple PT100 platinum resistance sensors within the furnace cavity, the mapping relationship between their respective measured values ​​and the temperatures of each region in the temperature field state vector is determined, forming a measurement mapping matrix H. This matrix H enables the system to effectively compare the actually measured observation vector Y(k) with the prior estimate of the temperature field state predicted by the model.

[0047] In alternative implementations, the mesh division of the temperature field region can be adjusted according to different types of commercial ovens (such as tunnel ovens or steam ovens), for example, by using finite element analysis for more detailed discretization. Historical test data used to calibrate the system matrix and control matrix B can also be updated periodically using an online adaptive identification method to address model parameter drift caused by aging of the oven cavity or changes in insulation performance. The selection of physical installation points for temperature sensors can be determined using an optimized layout algorithm to minimize the estimation error of the temperature field inversion.

[0048] In some embodiments, temperature field inversion includes initializing the estimated value of the temperature field state vector and its estimation error covariance based on the temperature field state model; aligning the time of multiple temperature data acquired within the same sampling period and performing filtering and denoising processing to obtain the observation vector at the current moment; using the optimal estimate from the previous moment, the actuator state data at the current moment, and the system matrix and control matrix in the temperature field state model to predict the prior estimate of the temperature field state at the current moment; comparing the observation vector at the current moment with the predicted observation value calculated from the prior estimate through the measurement mapping matrix, calculating the observation residual, and correcting the prior estimate based on Kalman gain calculation to obtain the optimal estimate at the current moment, i.e., the fused estimate of the temperature of each temperature field region in the furnace cavity.

[0049] The temperature field inversion module in this embodiment employs a Kalman filter algorithm (e.g., the unscented Kalman filter UKF) to overcome the influence of sensor measurement noise and model errors through dynamic fusion of model predictions and actual measurements, outputting reliable temperature estimates for each temperature field region within the furnace cavity. Upon system startup or reset, the temperature field state vector is first initialized with a high initial estimation error covariance to represent the uncertainty of the initial estimate. During each sampling period, multi-point temperature data collected by the multi-source sensor group are timestamped and digitally filtered to form the observation vector for the current moment.

[0050] In the specific inversion process, the first step is a time update (prediction) step: based on the optimal estimate of the previous moment, using the system matrix A and control matrix B in the temperature field state model, the prior estimate of the temperature field state at the current moment is predicted. Then, the measurement update (correction) step is performed: the observation vector is compared with the predicted observation value calculated through the measurement mapping matrix H, and the observation residual is calculated. The observation residual reflects the difference between the model prediction and the actual measurement. Kalman gain is calculated based on system noise and measurement noise. The Kalman gain calculation determines the correction weight for the prior estimate. Finally, the prior estimate is corrected using the observation residual and the Kalman gain calculation to obtain the optimal estimate for the current moment. This optimal estimate is the fused estimate of the temperature of each temperature field region in the furnace cavity.

[0051] In alternative or modified implementations, the temperature field inversion module can use an Extended Kalman Filter (EKF) or an Adaptive Kalman Filter, especially when the temperature field state model contains significant nonlinear heat conduction effects, where the EKF or Adaptive Kalman Filter can provide better estimation results. For filtering and denoising the observation vectors, median filtering or wavelet denoising methods can be used to improve the quality of the observation data. In specific application scenarios, the initialization process can incorporate historical operating data and employ Bayesian methods for more accurate initial state estimation.

[0052] In some embodiments, the door opening disturbance analysis includes real-time monitoring of door status data. When a door is detected to change from a closed state to an open state, or when the door opening degree exceeds a preset threshold, a door opening event is determined to have occurred, and the start time of the door opening is recorded. During the duration of the door opening event, based on the temperature field estimation results, the rate of decrease of the overall average temperature of the furnace cavity and / or the trend of the maximum temperature difference between each temperature field region are calculated in real time. Based on the duration of the door opening event, the door opening degree characteristics, the rate of decrease of the overall average temperature of the furnace cavity, and the ambient temperature, the amount of heat loss caused by the door opening is quantitatively calculated through a pre-established heat loss lookup table. At the same time, based on the trend of the maximum temperature difference, the intensity of cold air intrusion is assessed, which together constitute the door opening disturbance amount.

[0053] The door opening disturbance analysis module in this embodiment aims to accurately and in real-time quantify the thermodynamic disturbances caused by the opening operation of a commercial oven door, so as to provide accurate feedforward compensation basis for the staged compensation control module. Traditional control systems can only sense the hysteretic temperature drop after the door is opened, while this method can transform the external disturbance of door opening into a predictable amount of energy loss and temperature field non-uniformity. Real-time monitoring of door status data (such as mechanical limit switch signals or encoder readings) is the initial step in identifying door opening events.

[0054] In the specific analysis process, once a door opening event is determined, the door opening disturbance analysis module is immediately activated. During the door opening phase, the system continuously receives temperature field estimation results output by the temperature field inversion module. Based on these estimation results, the rate of decrease of the overall average temperature of the furnace cavity is calculated in real time, reflecting the rate of heat loss. Simultaneously, the system also calculates the temperature differences between the grids of each temperature field region to determine the trend of the maximum temperature difference. Subsequently, the system integrates parameters such as the door opening duration, door opening characteristics (if an encoder is used), ambient temperature, and the calculated temperature decrease rate to query a pre-calibrated heat loss lookup table. This lookup table, established through offline experiments, maps externally measurable parameters to heat loss (energy). Furthermore, the trend of the maximum temperature difference is used to assess the intensity of cold air intrusion, as the influx of cold air often leads to a sharp increase in local temperature differences. Ultimately, the heat loss and the intensity of cold air intrusion together constitute the door opening disturbance, which serves as the basis for phased compensation control.

[0055] In alternative or modified implementations, door status data can be obtained by using tilt sensor readings to determine the door's opening status and degree. The construction of the heat loss lookup table can further consider the oven's internal load (e.g., via a weighing sensor or preset load type parameters), incorporating the load's heat capacity effect into the heat loss calculation. The assessment of cold air intrusion intensity can be achieved not only by relying on the trend of maximum temperature difference changes but also by monitoring the rate of change of pressure difference within the oven cavity.

[0056] In some embodiments, the phased compensation control includes using the start time of the door opening event as the trigger point for entering the door opening phase and the moment when the door returns to the closed state as the trigger point for entering the door closing recovery phase. During the door opening phase, the output power of the heating actuator is limited to a level below the rated power, and the fan speed or damper opening is dynamically adjusted according to the real-time estimate of the door opening disturbance input to reduce heat loss due to forced convection. During the door closing recovery phase, the energy compensation requirement calculated based on the door opening disturbance is converted into feedforward control commands for one or more temperature control actuators. Simultaneously, based on the deviation between the temperature field estimation result and the set temperature control target, a feedback control command is generated through a feedback control algorithm. The feedforward control command generation and the feedback control command generation are superimposed and, after being constrained by the actuator's physical amplitude, the final temperature control quantity is generated.

[0057] The phased compensation control module in this embodiment is the core decision-maker of the entire temperature control system. Based on the input from the door status monitoring, it clearly divides the control strategy into an opening phase and a closing recovery phase, employing suppression control and feedforward-feedback superposition control respectively to cope with the dynamic disturbance of door opening.

[0058] During the door opening phase, the system immediately enters a suppression control state. To avoid the heating element continuously outputting high power when the door is open, resulting in a large amount of hot air being ejected or convection circulation exacerbating heat loss, the output power of the heating actuator is limited to a low maintenance level (e.g., maintaining only the minimum operating requirements predicted by the system temperature field state model). Simultaneously, based on the cold air intrusion intensity output by the door opening disturbance analysis module, the system dynamically adjusts the speed of the variable frequency fan or the damper opening to slow down air convection and heat exchange rates, reducing heat loss. When the door closes again, the system immediately switches to the door closing recovery phase. In this phase, based on the total heat loss quantified by the door opening disturbance analysis module, the system converts it into the required compensation energy and generates feedforward control commands. Feedforward control command generation provides a rapid, early energy replenishment. Simultaneously, based on the temperature field estimation results provided by the temperature field inversion module, the system calculates the deviation between the current temperature field state and the set target, and generates feedback control commands through traditional PID or other feedback control algorithms. Finally, the feedforward control command and the feedback control command are superimposed to form the final temperature control quantity, which is then sent to the temperature control actuator after being limited by the physical power of the actuator.

[0059] In alternative or modified implementations, the feedback control algorithm can employ model-based predictive control (MPC) or low-level QR control to handle the complexity of multiple-input multiple-output (MIMO) systems. Feedforward control command generation can incorporate compensation for oven wall thermal inertia in addition to being based on heat loss. During the door opening phase, besides reducing heating power, if the oven is equipped with a steam injection function, steam output can be temporarily stopped or reduced to prevent steam from accelerating heat loss when the door is opened.

[0060] In some embodiments, during the process of geometric discretization of the internal space of the furnace cavity, the specific way to divide the temperature field region is as follows: the furnace cavity is divided into three layers (upper, middle, and lower) along the height direction, into two rows (front and rear) along the depth direction, and into two sides (left and right) along the width direction. The furnace cavity is divided into at least six independent temperature field regions through the combination of three-dimensional space.

[0061] This implementation details the discretization method for the oven cavity in the temperature field state model module, aiming to ensure that the temperature field region mesh can effectively capture the temperature gradients and non-uniformities commonly found in commercial ovens through systematic three-dimensional partitioning. In actual circulating hot air ovens, the temperature typically exhibits significant gradients along the height direction (hot air buoyancy effect) and the depth direction (heat loss near the door).

[0062] In practice, the furnace cavity is discretized along its height into upper, middle, and lower layers to reflect vertical temperature stratification. Along its depth, it is discretized into a front row (near the door) and a rear row (near the fan / heating element) to reflect the difference in heat loss between the front and rear sections and the influence of the heating source. Through this combination of methods, the furnace cavity is divided into six main temperature field region grids (N=6): upper-front, upper-rear, middle-front, middle-rear, lower-front, and lower-rear. If the left and right sides in the width direction are further considered, it can be divided into twelve temperature field regions. This division ensures that the temperature field state vector can accurately describe the spatial temperature distribution within the furnace cavity, especially the front region sensitive to door opening and the vertical stratification susceptible to gravity.

[0063] In other alternative or modified implementations, if the oven is a tunnel oven (with continuous inlet and outlet), discretization can focus on segmentation along the length direction, while layering along the height direction can be simplified. For steam ovens, additional temperature field region meshes can be set near the steam injection nozzles to better model the exothermic effect of steam on local areas. In some simple oven structures, if the temperature field gradient is not significant, the partitioning can be simplified to two layers along the height direction or two rows along the depth direction, but N still needs to be greater than 2.

[0064] In some embodiments, the heat loss lookup table is constructed based on the following: the heat loss is positively correlated with the duration of door opening, positively correlated with the temperature difference inside and outside the furnace cavity, and positively correlated with the door opening degree; the intensity of cold air intrusion is positively correlated with the rate of change of the pressure difference inside and outside the furnace cavity and the trend of the maximum temperature difference.

[0065] This implementation details the physical basis and construction principles of the heat loss lookup table used to quantify door opening disturbances in the door opening disturbance analysis module, ensuring the accuracy and reliability of disturbance estimation. The heat loss lookup table is crucial for connecting measurable physical quantities with indirect heat loss energy.

[0066] During the construction of the lookup table, a large number of door-opening experiments were conducted under controlled conditions to systematically collect data under different operating conditions. Physical principles indicate that heat loss is mainly caused by convection and radiation. Specifically, the longer the door opening duration, the longer the heat exchange time, and the greater the heat loss; the greater the temperature difference between the inside and outside of the furnace cavity, the stronger the driving force for heat exchange, and the greater the heat loss; the larger the door opening, the larger the convection cross-section, and the greater the heat loss. These positive correlations are encoded into the lookup table's data structure, allowing the system to quickly obtain accurate heat loss based on real-time collected door opening duration, temperature difference, and opening characteristics through table lookup or interpolation methods during operation. Simultaneously, the intensity of cold air intrusion, as a major driving factor for temperature field inhomogeneity, is assessed based on the rate of change of the pressure difference between the inside and outside of the furnace cavity (reflecting the air exchange rate) and the trend of the maximum temperature difference (reflecting the degree of local temperature drop caused by the influx of cold air). A high rate of change of pressure difference and a significant trend of the maximum temperature difference indicate a strong cold air intrusion intensity.

[0067] In alternative or modified implementations, the heat loss lookup table can include more dimensions, such as the humidity or steam content inside the oven, as water vapor has a high specific heat capacity and its loss has a significant impact on the total heat loss. For assessing the intensity of cold air intrusion, an additional wind speed sensor can be introduced to directly measure the instantaneous airflow velocity at the oven cavity entrance. The data structure of the lookup table can be implemented using a multidimensional lookup table, a radial basis function (RBF)-based fitting model, or a lightweight neural network model to improve real-time calculation speed and accuracy.

[0068] In some embodiments, dynamically adjusting the fan speed or damper opening based on a real-time estimate of the door opening disturbance includes: based on a real-time estimated cold air intrusion intensity, when the intrusion intensity exceeds a limit, controlling the fan speed to decrease and adjusting the damper to reduce the opening of the furnace cavity and the outside environment to slow down the cold air exchange rate; when the intrusion intensity does not exceed a limit, maintaining the fan speed to keep the airflow in the furnace cavity stable and avoid aggravating temperature stratification.

[0069] During the dynamic adjustment process, the phased compensation control module receives the cold air intrusion intensity assessment results output by the door opening disturbance analysis module in real time. When the system determines that the cold air intrusion intensity exceeds the preset warning threshold (e.g., a sharp increase in the maximum temperature difference trend), it indicates that a large amount of external cold air is entering the oven cavity. At this time, the system immediately sends a command to the variable frequency fan in the temperature control actuator to reduce the speed, for example, to reduce the speed to 50% of the current operating speed, in order to weaken the forced convection intensity, thereby slowing down the speed at which cold air is drawn into the oven cavity and preventing the temperature stratification from worsening. At the same time, if the oven is equipped with a damper adjustment servo motor, the system will instruct the damper to reduce the opening of the cross-section connected to the outside, in order to further reduce the heat exchange rate. Conversely, if the cold air intrusion intensity does not exceed the limit, the system maintains the fan speed at the original setting to ensure the stability of the airflow and the uniform distribution of heat in the oven cavity, avoiding local stagnation due to excessive suppression.

[0070] This suppression control strategy, which dynamically adjusts the fan speed and damper opening, effectively balances the contradiction between "reducing heat loss" and "maintaining temperature field uniformity." By suppressing the intrusion of cold air only when the risk is high, it minimizes the adverse effects of the door opening stage on the airflow stability within the furnace cavity, thus creating favorable conditions for rapid and uniform heating during the door closing and recovery stage.

[0071] In alternative or modified implementations, the fan speed adjustment can employ a tiered strategy, for example, dividing it into low, medium, and high levels based on the intensity of cold air intrusion, corresponding to different speed suppression ratios. The damper adjustment can also be combined with exhaust valve control, temporarily closing the exhaust valve during the door opening phase to reduce heat loss through the exhaust passage. In some ovens that rely on natural convection, this step can be simplified to simply adjusting the damper opening.

[0072] In some embodiments, the energy compensation requirement calculated based on the door opening disturbance is converted into feedforward control commands for one or more temperature control actuators, including: converting the quantified heat loss into the total energy to be supplemented; distributing the total energy non-uniformly to different heating execution loops according to the temperature drop difference of each temperature field grid during the door opening stage, increasing the feedforward power ratio of the heaters corresponding to the temperature field areas where the temperature drop exceeds the limit, and increasing the feedforward command of the fan speed to enhance the stirring and mixing of hot air in the furnace cavity.

[0073] This embodiment details the core innovation of the phased compensation control module in generating feedforward control commands during the door-closing recovery phase: non-uniform feedforward compensation. This feature aims to address the temperature field non-uniformity caused by door-opening disturbances by specifically compensating for the area with the largest temperature drop, thereby achieving rapid and balanced temperature field recovery.

[0074] In the specific conversion process, the heat loss provided by the door opening disturbance analysis module is converted into a total energy that needs to be replenished. This total energy is the basis for feedforward control. Unlike traditional schemes that uniformly distribute the total energy to all heating elements, this method, based on the temperature field estimation results output by the temperature field inversion module, determines which temperature field regions (e.g., the lower front region) experience a temperature drop exceeding the average level during the door opening phase. Subsequently, the system implements non-uniform distribution: the majority (e.g., 55%) of the total energy is concentrated on the heating execution loops corresponding to the temperature field regions where the temperature drop exceeds the limit, with the remainder distributed to other regions. For example, if the oven has independently controlled upper, lower, and rear heaters, more feedforward power is allocated to the heaters at the corresponding lower temperatures. Simultaneously, to accelerate heat mixing and transfer, the system also generates a feedforward command to increase the fan speed (e.g., by 10%). This command, independent of the heating power distribution, is used to enhance the stirring and mixing of hot air within the oven cavity.

[0075] In alternative or modified implementations, the non-uniformity ratio of energy distribution can be dynamically calculated based on the oven's physical structure and the thermal coupling matrix of the temperature field grid, rather than using a fixed ratio. Feedforward control commands can be generated not only for heating actuators but also for other temperature control actuators (such as steam jets or dampers) to handle more complex baking environments. The fan speed feedforward command can be further refined into a dynamic programming strategy of high speed during the initial recovery phase and reduced speed during the later recovery phase.

[0076] Figure 3 A schematic diagram of a commercial oven precision temperature control system based on multi-sensor fusion is provided as an embodiment of this application, as shown below. Figure 3 As shown, a commercial oven precision temperature control system 300 based on multi-sensor fusion in this embodiment includes: a data acquisition module 301, a temperature field modeling module 302, a temperature field inversion module 303, a door opening disturbance analysis module 304, a phased compensation control module 305, and an execution control module 306.

[0077] Data acquisition module 301 is used to acquire multi-source sensor data of commercial ovens, including multi-point temperature data of oven cavity, actuator status data and door status data; The temperature field modeling module 302 is used to divide the oven cavity of the commercial oven into multiple temperature field regions, establish a temperature field state model characterizing the temperature of each temperature field region, and establish a measurement mapping relationship between the multi-source sensing data and the temperature field state model. The temperature field inversion module 303 is used to perform fusion estimation on the multi-source sensing data based on the temperature field state model and the measurement mapping relationship to obtain the temperature field estimation result of the furnace cavity; The door opening disturbance analysis module 304 is used to identify door opening events based on the door status data and estimate the door opening disturbance amount based on the temperature field estimation results. The phased compensation control module 305 is used to switch the temperature control process into an opening stage and a closing recovery stage based on the door opening event. In the opening stage, the heating output is suppressed and controlled according to the door opening disturbance. In the closing recovery stage, a feedforward compensation control quantity is generated based on the door opening disturbance and superimposed with the feedback control quantity generated based on the temperature field estimation result to obtain the temperature control quantity. The execution control module 306 is used to control the temperature control actuator of the commercial oven according to the temperature control quantity, so that the oven cavity temperature field meets the set temperature control target.

[0078] Optionally, the temperature field modeling module 302 is specifically used for: Based on the physical structure of the oven cavity of a commercial oven, the internal space of the oven cavity is geometrically discretized into N interrelated temperature field regions, where N is an integer greater than 2; A temperature field state vector is defined to characterize the overall temperature distribution of the furnace cavity at the current moment. The temperature field state vector is composed of N equivalent temperature values ​​of the temperature field regions arranged sequentially. By using historical test data of the oven, a system matrix describing the thermal coupling relationship between each temperature field region and a control matrix describing the influence of the temperature control actuator on the temperature of each temperature field region were calibrated and obtained. Based on the physical installation positions of all temperature sensors in the furnace cavity, the mapping weight between the measured value of each temperature sensor and the temperature of one or more of the temperature field regions is determined, forming a measurement mapping matrix, and completing the establishment of the mapping relationship from the temperature field state vector to the actual collectable multi-point temperature data.

[0079] Optionally, the temperature field inversion module 303 is specifically used for: Based on the temperature field state model, initialize the estimated value of the temperature field state vector and its estimation error covariance; The multi-point temperature data acquired within the same sampling period are time-aligned and filtered and denoised to obtain the observation vector at the current moment. Using the temperature field estimation results from the previous moment, the actuator state data at the current moment, and the system matrix and control matrix in the temperature field state model, the prior estimate of the temperature field state at the current moment is predicted. The observation vector at the current moment is compared with the predicted observation value calculated by the prior estimate through the measurement mapping matrix. The observation residual is calculated, and the prior estimate is corrected based on the Kalman gain to obtain the optimal temperature field estimation result at the current moment, that is, the fused estimate of the temperature of each temperature field region in the furnace cavity.

[0080] Optionally, the door opening disturbance analysis module 304 is specifically used for: The door status data is monitored in real time. When the door changes from closed to open or the door opening degree exceeds a preset threshold, an opening event is determined to have occurred, and the opening start time is recorded. During the duration of the door opening event, based on the temperature field estimation results, the rate of decrease of the overall average temperature of the furnace cavity and / or the trend of the maximum temperature difference between each temperature field region are calculated in real time. Based on the duration of the door opening event, the door opening characteristics, the rate of decrease of the overall average temperature of the furnace cavity, and the ambient temperature, the heat loss caused by the door opening is quantitatively calculated using a pre-established heat loss lookup table. At the same time, based on the trend of the maximum temperature difference change, the intensity of cold air intrusion is assessed, which together constitute the door opening disturbance.

[0081] Optionally, the phased compensation control module 305 is specifically used for: The start time of the door opening event is taken as the trigger point for entering the door opening stage, and the moment when the door returns to the closed state is taken as the trigger point for entering the door closing and recovery stage. During the door opening phase, the output power of the heating actuator is limited to a level below the rated power, and the fan speed or damper opening is dynamically adjusted according to the real-time estimate of the door opening disturbance to reduce heat loss due to forced convection. During the door closing recovery phase, the energy compensation requirement calculated based on the door opening disturbance is converted into feedforward control commands for one or more temperature control actuators. At the same time, based on the deviation between the temperature field estimation result and the set temperature control target, feedback control commands are generated through a feedback control algorithm. The feedforward control command and the feedback control command are superimposed and then subjected to physical limiting constraints by the actuator to generate the final temperature control quantity.

[0082] Optionally, in the temperature field modeling module 302, during the process of geometric discretization of the internal space of the furnace cavity, the specific method for dividing the temperature field region is as follows: along the height direction of the furnace cavity, it is divided into three layers: upper, middle and lower; along the depth direction of the furnace cavity, it is divided into two rows: front and back; and along the width direction of the furnace cavity, it is divided into left and right sides. Through the combination of three-dimensional space, the furnace cavity is divided into at least six independent temperature field regions.

[0083] Optionally, in the door opening disturbance analysis module 304, the heat loss lookup table is constructed based on the following: the heat loss is positively correlated with the door opening duration, positively correlated with the temperature difference inside and outside the furnace cavity, and positively correlated with the door opening degree; the cold air intrusion intensity is positively correlated with the rate of change of the pressure difference inside and outside the furnace cavity and the trend of the maximum temperature difference.

[0084] Optionally, when the phased compensation control module 305 dynamically adjusts the fan speed or damper opening based on the real-time estimated value of the door opening disturbance, it is specifically used for: Based on the real-time estimated cold air intrusion intensity, when the intrusion intensity exceeds the limit, the fan speed is reduced and the damper is adjusted to reduce the opening of the furnace cavity and the outside, so as to slow down the cold air exchange rate. When the intrusion intensity does not exceed the limit, the fan speed is maintained to keep the airflow in the furnace cavity stable and avoid aggravating temperature stratification.

[0085] Optionally, when the phased compensation control module 305 converts the energy compensation requirement calculated based on the door opening disturbance into feedforward control commands for one or more temperature control actuators, it is specifically used for: Based on the quantified heat loss, it is equivalent to the total energy that needs to be replenished; Based on the difference in temperature drop in each temperature field region during the opening stage, the total energy is non-uniformly distributed to different heating execution circuits. This increases the feedforward power ratio of the heaters corresponding to temperature field regions where the temperature drop exceeds the limit, while simultaneously increasing the feedforward command of the fan speed to enhance the stirring and mixing of hot air in the furnace cavity.

[0086] The system in this embodiment can be used to execute the methods of any of the above embodiments, and its implementation principle and technical effect are similar, so they will not be described again here.

Claims

1. A method for precise temperature control of a commercial oven based on multi-sensor fusion, characterized in that, include: Data acquisition: Acquire multi-source sensor data of the commercial oven, including multi-point temperature data of the oven cavity, actuator status data, and door status data; Temperature field modeling: The oven cavity of the commercial oven is divided into multiple temperature field regions, a temperature field state model characterizing the temperature of each temperature field region is established, and a measurement mapping relationship between the multi-source sensor data and the temperature field state model is established. Temperature field inversion: Based on the temperature field state model and the measurement mapping relationship, the multi-source sensor data is fused and estimated to obtain the temperature field estimation result of the furnace cavity; Door opening disturbance analysis: Identify door opening events based on the door status data, and estimate the door opening disturbance amount based on the temperature field estimation results; Phased compensation control: Based on the door opening event, the temperature control process is switched to a door opening stage and a door closing recovery stage. In the door opening stage, the heating output is suppressed and controlled according to the door opening disturbance. In the door closing recovery stage, a feedforward compensation control quantity is generated based on the door opening disturbance and superimposed with the feedback control quantity generated based on the temperature field estimation result to obtain the temperature control quantity. Execution control: Control the temperature control actuator of the commercial oven according to the temperature control quantity, so that the oven cavity temperature field meets the set temperature control target.

2. The method according to claim 1, characterized in that, The temperature field modeling includes: Based on the physical structure of the oven cavity of a commercial oven, the internal space of the oven cavity is geometrically discretized into N interrelated temperature field regions, where N is an integer greater than 2; A temperature field state vector is defined to characterize the overall temperature distribution of the furnace cavity at the current moment. The temperature field state vector is composed of N equivalent temperature values ​​of the temperature field regions arranged sequentially. By using historical test data of the oven, a system matrix describing the thermal coupling relationship between each temperature field region and a control matrix describing the influence of the temperature control actuator on the temperature of each temperature field region were calibrated and obtained. Based on the physical installation positions of all temperature sensors in the furnace cavity, the mapping weight between the measured value of each temperature sensor and the temperature of one or more of the temperature field regions is determined, forming a measurement mapping matrix, and completing the establishment of the mapping relationship from the temperature field state vector to the actual collectable multi-point temperature data.

3. The method according to claim 2, characterized in that, The temperature field inversion includes: Based on the temperature field state model, initialize the estimated value of the temperature field state vector and its estimation error covariance; The multi-point temperature data acquired within the same sampling period are time-aligned and filtered and denoised to obtain the observation vector at the current moment. Using the temperature field estimation results from the previous moment, the actuator state data at the current moment, and the system matrix and control matrix in the temperature field state model, the prior estimate of the temperature field state at the current moment is predicted. The observation vector at the current moment is compared with the predicted observation value calculated by the prior estimate through the measurement mapping matrix. The observation residual is calculated, and the prior estimate is corrected based on the Kalman gain to obtain the optimal temperature field estimation result at the current moment, that is, the fused estimate of the temperature of each temperature field region in the furnace cavity.

4. The method according to claim 2, characterized in that, The door opening disturbance analysis includes: The door status data is monitored in real time. When the door changes from closed to open or the door opening degree exceeds a preset threshold, an opening event is determined to have occurred, and the opening start time is recorded. During the duration of the door opening event, based on the temperature field estimation results, the rate of decrease of the overall average temperature of the furnace cavity and / or the trend of the maximum temperature difference between each temperature field region are calculated in real time. Based on the duration of the door opening event, the door opening characteristics, the rate of decrease of the overall average temperature of the furnace cavity, and the ambient temperature, the heat loss caused by the door opening is quantitatively calculated using a pre-established heat loss lookup table. At the same time, based on the trend of the maximum temperature difference change, the intensity of cold air intrusion is assessed, which together constitute the door opening disturbance.

5. The method according to claim 4, characterized in that, The phased compensation control includes: The start time of the door opening event is taken as the trigger point for entering the door opening stage, and the moment when the door returns to the closed state is taken as the trigger point for entering the door closing and recovery stage. During the door opening phase, the output power of the heating actuator is limited to a level below the rated power, and the fan speed or damper opening is dynamically adjusted according to the real-time estimate of the door opening disturbance to reduce heat loss due to forced convection. During the door closing recovery phase, the energy compensation requirement calculated based on the door opening disturbance is converted into feedforward control commands for one or more temperature control actuators. At the same time, based on the deviation between the temperature field estimation result and the set temperature control target, feedback control commands are generated through a feedback control algorithm. The feedforward control command and the feedback control command are superimposed and then subjected to physical limiting constraints by the actuator to generate the final temperature control quantity.

6. The method according to claim 2, characterized in that, During the process of geometric discretization of the internal space of the furnace cavity, the specific method for dividing the temperature field region is as follows: the furnace cavity is divided into three layers (upper, middle, and lower) along the height direction, two rows (front and rear) along the depth direction, and two sides (left and right) along the width direction. The furnace cavity is divided into at least six independent temperature field regions through the combination of three-dimensional space.

7. The method according to claim 4, characterized in that, The heat loss lookup table is constructed based on the following: the heat loss is positively correlated with the duration of door opening, positively correlated with the temperature difference inside and outside the furnace cavity, and positively correlated with the door opening degree; the intensity of cold air intrusion is positively correlated with the rate of change of the pressure difference inside and outside the furnace cavity and the trend of the maximum temperature difference.

8. The method according to claim 5, characterized in that, The step of dynamically adjusting the fan speed or damper opening based on the real-time estimate of the door opening disturbance includes: Based on the real-time estimated cold air intrusion intensity, when the intrusion intensity exceeds the limit, the fan speed is reduced and the damper is adjusted to reduce the opening of the furnace cavity and the outside, so as to slow down the cold air exchange rate. When the intrusion intensity does not exceed the limit, the fan speed is maintained to keep the airflow in the furnace cavity stable and avoid aggravating temperature stratification.

9. The method according to claim 5, characterized in that, The step of converting the energy compensation requirement calculated based on the door opening disturbance into feedforward control commands for one or more temperature control actuators includes: Based on the quantified heat loss, it is equivalent to the total energy that needs to be replenished; Based on the difference in temperature drop in each temperature field region during the opening stage, the total energy is non-uniformly distributed to different heating execution circuits. This increases the feedforward power ratio of the heaters corresponding to temperature field regions where the temperature drop exceeds the limit, while simultaneously increasing the feedforward command of the fan speed to enhance the stirring and mixing of hot air in the furnace cavity.

10. A commercial oven precision temperature control system based on multi-sensor fusion, characterized in that, include: The data acquisition module is used to acquire multi-source sensor data of the commercial oven, including multi-point temperature data of the oven cavity, actuator status data and door status data. The temperature field modeling module is used to divide the oven cavity of the commercial oven into multiple temperature field regions, establish a temperature field state model characterizing the temperature of each temperature field region, and establish a measurement mapping relationship between the multi-source sensor data and the temperature field state model. The temperature field inversion module is used to perform fusion estimation on the multi-source sensor data based on the temperature field state model and the measurement mapping relationship to obtain the temperature field estimation result of the furnace cavity; The door opening disturbance analysis module is used to identify door opening events based on the door status data and estimate the door opening disturbance amount based on the temperature field estimation results. The phased compensation control module is used to switch the temperature control process into an opening stage and a closing recovery stage based on the door opening event. In the opening stage, the heating output is suppressed and controlled according to the door opening disturbance. In the closing recovery stage, a feedforward compensation control quantity is generated based on the door opening disturbance and superimposed with the feedback control quantity generated based on the temperature field estimation result to obtain the temperature control quantity. The execution control module is used to control the temperature control actuator of the commercial oven according to the temperature control quantity, so that the oven cavity temperature field meets the set temperature control target.

Images