Intelligent toaster temperature compensation method and system linked with Internet of Things

Abstract

The invention relates to the technical field of intelligent household electrical appliance control, and discloses an Internet of Things linked toaster intelligent temperature compensation method and system. The method comprises the steps that an intelligent temperature compensation algorithm model is formed through historical baking data training, and the intelligent temperature compensation algorithm model at least comprises an environment temperature influence factor calculation unit, a cavity aging coefficient calculation unit and a bread category baking curve database; the environment temperature influence factor calculation unit is used for calculating an environment influence weight in combination with the real-time environment parameters obtained by the environment perception linkage module, and the cavity aging coefficient calculation unit is used for calculating a cavity aging coefficient according to the historical use duration of the toaster; and generating a temperature compensation value at the current moment based on the real-time data corresponding to the toaster, the environmental influence weight, the cavity aging coefficient and the standard baking curve corresponding to the bread category. According to the invention, breakthrough improvement of a traditional toaster temperature control technology is realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent household appliance control, and in particular to a smart temperature compensation method and system for a toaster connected with the Internet of Things. BACKGROUND

[0002] The temperature control scheme of the current market toasters mainly relies on single-point feedback of the built-in sensor of the equipment, and can only adjust the heating power according to the preset fixed program, which has the following significant defects: 1. Insufficient environmental adaptability: the existing technology does not consider the influence of environmental temperature and humidity on the heating efficiency of the cavity, and when the environmental temperature and humidity change (such as low temperature or humid environment in winter), the actual toasting temperature deviates greatly from the target temperature, resulting in unstable bread toasting effect.

[0003] 2. Lack of device aging compensation: after long-term use, the performance of the heating element of the toaster and the loss of the cavity insulation layer will cause the heat efficiency to decrease, but the traditional scheme lacks quantitative evaluation and dynamic compensation mechanism for the aging state of the equipment, and cannot maintain long-term stable toasting quality.

[0004] 3. Single control strategy: the existing control algorithm only calls a fixed toasting curve based on the type of bread, does not combine dynamic parameters such as real-time toasting stage progress and the number of bread placed, and does not build a multi-dimensional correlation model of environmental parameters-equipment state-toasting demand, making it difficult to achieve fine temperature adjustment.

[0005] 4. Lack of Internet of Things connection: the traditional toaster is an independent device, which cannot interact with environmental sensors, smart meters and other peripheral Internet of Things devices, and cannot be optimized through a cloud server, so the degree of intelligence is low.

[0006] Although there are a few schemes related to temperature compensation of household appliances in the prior art, they are all limited to local sensor feedback or simple rule control, and have never disclosed the combination of Internet of Things connection, real-time sensing of environmental parameters, calculation of device aging coefficient and machine learning algorithm, to build an intelligent system architecture containing a cloud dynamic compensation model and a closed-loop feedback mechanism. SUMMARY

[0007] The present application provides a smart temperature compensation method and system for a toaster connected with the Internet of Things, which aims to solve the problem that the prior art is limited to local sensor feedback or simple rule control, and has never disclosed the combination of Internet of Things connection, real-time sensing of environmental parameters, calculation of device aging coefficient and machine learning algorithm, to build an intelligent system architecture containing a cloud dynamic compensation model and a closed-loop feedback mechanism.

[0008] In a first aspect, the present application provides a smart temperature compensation system for a toaster connected with the Internet of Things, comprising: The data acquisition module is used for acquiring real-time data of the toaster; the real-time data includes collected baking cavity temperature, bread placement quantity, bread type parameters and current baking stage progress; The Internet of Things communication module is used for sending the real-time data and device identification to the computer device and receiving control instructions issued by the computer device; The heating control module is used for adjusting the corresponding power output or heating time of the heating element of the toaster according to the control instructions; The environment perception linkage module is configured to acquire real-time environment temperature and humidity data of the environment where the toaster is located, and support data interaction with peripheral Internet of Things devices such as smart meters and environment temperature and humidity sensors. The computer device is internally provided with an intelligent temperature compensation algorithm model formed by training historical baking data, the intelligent temperature compensation algorithm model at least includes an environment temperature influence factor calculation unit, a cavity aging coefficient calculation unit and a bread category baking curve database, the environment temperature influence factor calculation unit is used for calculating an environment influence weight in combination with real-time environment parameters acquired by the environment perception linkage module, and the cavity aging coefficient calculation unit is used for calculating a cavity aging coefficient according to historical use time of the toaster; based on the real-time data, the environment influence weight, the cavity aging coefficient and the standard baking curve of the corresponding bread category, a temperature compensation value at the current time is generated.

[0009] In some embodiments, after the temperature compensation value at the current time is generated, the method further includes: performing deviation analysis on the compensated cavity temperature data fed back by the toaster and the standard baking curve through a dynamic time warping algorithm, and real-time correcting the temperature compensation value and updating compensation parameters of the algorithm model.

[0010] In some embodiments, the intelligent temperature compensation algorithm model is constructed by using a machine learning algorithm, can optimize the environment parameter influence weight and the aging coefficient correction rule through continuous learning of new baking data, and the toaster terminal device, the environment perception linkage module and the computer device realize data interaction and linkage control through an Internet of Things platform.

[0011] In some embodiments, the intelligent temperature compensation algorithm model formed by training historical baking data includes: multi-dimensional feature labeling of historical collected baking cavity temperature data, corresponding environment temperature and humidity data, toaster use time data, bread type and baking result data, iterative training of parameters of the environment temperature influence factor calculation unit, the cavity aging coefficient calculation unit and the bread category baking curve database by using a gradient descent optimization algorithm, and establishment of the intelligent temperature compensation algorithm model.

[0012] In some embodiments, the environment influence weight is calculated according to the real-time environment parameters obtained by the environment perception linkage module, including: according to the difference interval of the real-time environment temperature and the standard reference temperature, the influence coefficient of the environment humidity on the heat conduction efficiency, and through the preset environment parameter influence rule base, a corresponding weight adjustment factor is matched, the weight adjustment factor is associated with the historical baking temperature deviation data under the same environment condition, a dynamic adjustable environment temperature influence weight calculation mechanism for the cavity heating efficiency is formed, and the environment influence weight is calculated.

[0013] In some embodiments, the cavity aging coefficient is calculated according to the toasters historical use time, including: an aging evaluation model based on the cumulative working time of the heating element, the loss degree of the cavity insulation layer and the temperature sensor drift amount is established, the slope difference between the standard heating curve of the new device and the actual heating curve of the current device is compared, the cavity thermal efficiency attenuation curve is fitted combined with the use time, and the cavity aging coefficient reflecting the degradation degree of the cavity heating performance is generated.

[0014] In some embodiments, the temperature compensation value at the current moment is generated based on the real-time data, the environment influence weight, the cavity aging coefficient and the standard baking curve of the corresponding bread type, including: according to the bread type parameter, a target temperature curve in the bread type baking curve database is matched and called, and the target temperature curve is corrected in terms of environment temperature by the environment influence weight for the current baking stage progress; the thermal efficiency compensation adjustment of the corrected target temperature is made based on the cavity aging coefficient, and the real-time dynamic temperature compensation value is formed.

[0015] In a second aspect, the present application provides a smart temperature compensation method of Internet of Things linkage toaster, applied to the computer equipment of the smart temperature compensation system of Internet of Things linkage toaster provided in any embodiment of the present application; the method comprises: The intelligent temperature compensation algorithm model formed by training the historical baking data, the intelligent temperature compensation algorithm model at least includes an environment temperature influence factor calculation unit, a cavity aging coefficient calculation unit and a bread type baking curve database, the environment temperature influence factor calculation unit is used to calculate the environment influence weight combined with the real-time environment parameters obtained by the environment perception linkage module, and the cavity aging coefficient calculation unit is used to calculate the cavity aging coefficient according to the historical use time of the toaster; Based on the real-time data, the environment influence weight, the cavity aging coefficient and the standard baking curve of the corresponding bread type of the toaster, the temperature compensation value at the current moment is generated.

[0016] In a third aspect, the present application provides a computer equipment, the computer equipment includes a memory and a processor; the memory is used to store a computer program; the processor is used to execute the computer program and realize the method provided in any embodiment of the present application when the computer program is executed.

[0017] In a fourth aspect, the present application provides a computer readable storage medium, which stores a computer program, and the computer readable instructions are executed by the processor to make one or more processors execute the method provided by any of the embodiments of the present application.

[0018] The present application breaks through the technical bottleneck of traditional single-point control by introducing an environment perception linkage module, a cloud intelligent algorithm model and multi-dimensional data interaction. By collecting data such as cavity temperature, environmental parameters, equipment aging state and bread type in real time, an intelligent algorithm model containing environmental influence factors and cavity aging coefficient is constructed to realize dynamic correction of the baking temperature, solve the temperature deviation problem caused by environmental changes and equipment aging in traditional schemes, and significantly improve the baking uniformity and product consistency. Through real-time data interaction between computer equipment (such as a cloud server) and a toaster, the dynamic time warping algorithm is used to analyze the deviation between the actual baking curve and the standard curve to form a closed-loop control of “collection-compensation-correction”, so that the system can adapt to different use scenarios and continuously optimize the compensation strategy, avoiding the rigid defects of fixed program control. Support data interaction with environmental temperature and humidity sensors, smart meters and other devices to build an intelligent scene with multi-device collaboration, and continuously learn new data through cloud machine learning algorithm to improve the system's adaptability to complex environments and promote the upgrading of traditional home appliances to intelligence and networking. The aging evaluation model quantifies the degree of cavity thermal efficiency decay, and dynamically adjusts the compensation parameters combined with historical use data to effectively alleviate the impact of equipment aging on baking quality, prolong the efficient service life of the toaster, and reduce the user's use cost.

[0019] It should be understood that the above general description and the following detailed description are only exemplary and explanatory, and cannot limit the present application. BRIEF DESCRIPTION OF DRAWINGS

[0020] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are some embodiments of the present application, and other drawings can also be obtained by those skilled in the art without creative labor.

[0021] Figure 1 is a structural schematic block diagram of the Internet of Things linkage toaster intelligent temperature compensation system provided by an embodiment of the present application; Figure 2 is a step schematic flow chart of the Internet of Things linkage toaster intelligent temperature compensation method provided by an embodiment of the present application; Figure 3 is a structural schematic block diagram of a computer device provided by an embodiment of the present application.

[0022] It should be understood that the above general description and the following detailed description are only exemplary and explanatory and are not restrictive of the application. DETAILED DESCRIPTION

[0023] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work fall within the scope of protection of the present application.

[0024] The flowcharts shown in the drawings are only exemplary and do not necessarily include all the contents and operations / steps, nor are they necessarily executed in the described order. For example, some operations / steps can be further decomposed, combined or partially merged, so that the actual execution order can be changed according to the actual situation.

[0025] It should be understood that, in order to clearly describe the technical solutions of the embodiments of the present application, in the embodiments of the present application, the terms "first", "second", etc. are used to distinguish the same items or similar items with basically the same functions and effects. Those skilled in the art can understand that the terms "first", "second", etc. do not limit the quantity and execution order, and the terms "first", "second", etc. also do not necessarily mean that they are different.

[0026] It should be understood that the terms used in this application specification are only for the purpose of describing specific embodiments and do not intend to limit the present application. As used in the present application specification and the appended claims, unless otherwise clear from the context, the singular forms "a", "an" and "the" are intended to include the plural forms.

[0027] It should also be understood that the term "and / or" used in the present application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes these combinations.

[0028] Some embodiments of the present application will be described in detail below with reference to the accompanying drawings. The following embodiments and features in the embodiments can be combined with each other without conflict.

[0029] The toasters temperature control scheme in the current market mainly relies on single-point feedback of the built-in sensor in the equipment, and can only adjust the heating power according to the preset fixed program, which has the following significant defects: 1. Insufficient environmental adaptability: The existing technology does not consider the influence of environmental temperature and humidity on the heating efficiency of the cavity. When the environmental temperature and humidity change (such as low temperature or humid environment in winter), the actual baking temperature deviates greatly from the target temperature, resulting in unstable bread baking effect.

[0030] 2. Lack of device aging compensation: After long-term use, factors such as performance degradation of toaster heating elements and loss of cavity insulation layer will cause heat efficiency to decline. However, the traditional scheme lacks quantitative evaluation and dynamic compensation mechanism for the aging state of the device, and cannot maintain long-term stable baking quality.

[0031] 3. Single control strategy: The existing control algorithm only calls fixed baking curves based on bread type, without combining dynamic parameters such as real-time baking stage progress and bread placement quantity, and without building a multi-dimensional correlation model of environmental parameters, device state and baking demand, making it difficult to achieve fine temperature regulation.

[0032] 4. Lack of Internet of Things linkage: Traditional toasters are independent devices that cannot interact with environmental sensors, smart meters and other peripheral Internet of Things devices, and cannot perform algorithm iteration optimization through cloud servers, resulting in low intelligence level.

[0033] Although there are a few schemes in the existing technology that involve temperature compensation of household appliances, they are all limited to local sensor feedback or simple rule control, and have never disclosed the combination of Internet of Things linkage, real-time sensing of environmental parameters, calculation of device aging coefficient and machine learning algorithm, to build an intelligent system architecture containing a cloud dynamic compensation model and a closed-loop feedback mechanism.

[0034] To solve the above problems, please refer to Figure 1The application provides a smart temperature compensation system for a toaster connected with the Internet of Things, comprising: a data acquisition module for acquiring real-time data of the toaster; the real-time data comprises the acquisition of the baking cavity temperature, the number of bread placed, the bread type parameters and the current baking stage progress; an Internet of Things communication module for sending the real-time data and the device identification to a computer device and receiving the control instructions issued by the computer device; a heating control module for adjusting the corresponding power output or heating time of the heating element of the toaster according to the control instructions; an environment perception linkage module configured to acquire real-time environmental temperature and humidity data of the environment where the toaster is located, and support data interaction with peripheral Internet of Things devices such as smart meters and environmental temperature and humidity sensors; a computer device internally provided with an intelligent temperature compensation algorithm model formed by training historical baking data, wherein the intelligent temperature compensation algorithm model at least comprises an environmental temperature influence factor calculation unit, a cavity aging coefficient calculation unit and a bread category baking curve database, the environmental temperature influence factor calculation unit is used to calculate the environmental influence weight in combination with the real-time environmental parameters acquired by the environment perception linkage module, and the cavity aging coefficient calculation unit is used to calculate the cavity aging coefficient according to the historical use time of the toaster; based on the real-time data, the environmental influence weight, the cavity aging coefficient and the standard baking curve of the corresponding bread category, a temperature compensation value at the current time is generated.

[0035] Specifically, the system constructs a multi-dimensional intelligent architecture of "data acquisition-cloud computing-intelligent control-closed loop feedback", which contains five core modules and realizes the full-factor linkage of environment, equipment and baking demand. The data acquisition module acquires the baking cavity temperature (accuracy ±0.5℃) in real time through the built-in temperature sensor, deploys the pressure sensor or the weight sensing device to detect the number of bread placed (supports 1-4 pieces recognition). The user input acquires the bread type parameters (such as white bread, whole wheat bread, baguette, etc., corresponding to different baking curves) through the device panel or the matching APP. The baking stage (preheating, constant baking, coloring stage) is divided based on the heating time, temperature change rate and other parameters, and the current progress is marked in real time.

[0036] The Internet of Things communication module uploads the device identification (unique ID) and real-time data to the cloud server through Wi-Fi, Bluetooth or Zigbee protocol, and receives the control instructions issued by the cloud (delay <500ms). The AES-128 encryption algorithm is adopted to ensure the transmission safety and prevent data tampering.

[0037] The environmental perception linkage module obtains the environmental temperature (-10°C~40°C) and humidity (20%~90%RH) in real time through the built-in or externally connected temperature and humidity sensor (accuracy ±2°C, ±5%RH), supports interaction with the smart meter to obtain real-time electricity price and power limit data (for energy-saving mode), and interworks with surrounding devices (such as kitchen air conditioner and dehumidifier) through the Internet of Things platform (such as Home Assistant and Ali Cloud IoT) to build a kitchen micro-environment model.

[0038] The heating control module supports PWM (pulse width modulation) adjustment of the heating element power (range 0~100%) or dynamic adjustment of the heating time length (accuracy ±1 second), and is compatible with multiple heating modes such as quartz tube and infrared heating. The temperature upper limit (such as 250°C) and timeout protection (such as 30 minutes of automatic power-off) are set.

[0039] The computer device (cloud + edge coordination) corresponding to the intelligent temperature compensation algorithm model includes: an environmental temperature influence factor calculation unit: based on historical data, a regression model is trained, the real-time environmental temperature (T_env) and humidity (H_env) are input, and the environmental influence weight coefficient K_env is output (for example, in a low-temperature and high-humidity environment, K_env=1.2, the heating power is compensated). A cavity aging coefficient calculation unit: through the device use time (t), the heating number (n) and the historical failure data, an aging decay model (such as an exponential decay function α=0.99^t+0.01n) is constructed to quantify the degree of heat efficiency decline. A bread type baking curve database: stores the standard temperature-time curve of different bread types (such as white bread which needs 180°C±10°C for 3 minutes to color), and supports user-defined curve import. Compensation value generation logic: T_compensate=T_target×(1+K_env×α)+ΔT_phase; wherein ΔT_phase is the dynamic correction value of the current baking phase (such as an additional +5°C in the coloring phase).

[0040] The environmental parameters (T_env, H_env), device state (aging coefficient α) and baking demand (bread type, quantity) are input into the machine learning model (such as random forest or LSTM) to build an “environment-device-demand” correlation matrix, which replaces the traditional single preset program. Through real-time feedback of the cavity temperature, the compensation value is updated every 10 seconds, forming a closed loop of “collection-computation-control-re-collection”, and the error is controlled within ±3°C.

[0041] When the cloud server is used as the computer device, it can aggregate data of multiple devices, continuously optimize model parameters (updated automatically once a week), and solve the problem of insufficient data of a single machine; the edge node (local device) stores the basic model to ensure that it can still be compensated based on historical data when offline.

[0042] Sensor configuration: NTC temperature sensor (Model MF52B) built-in the cavity, SHT30 temperature and humidity module deployed externally, and a thin film pressure sensor (distributed at the bottom of the tray) for bread quantity detection. Communication module: integrated ESP32-WROOM-32 chip, supporting Wi-Fi 2.4G and BLE 5.0, communicating with the cloud through the MQTT protocol. Heating element: nickel-chromium alloy heating tube with adjustable power, cooperating with a reflective plate to evenly distribute heat, with a power adjustment resolution of 10W.

[0043] Data acquisition and upload (device end) includes: initialization: user selects the bread type (input or scans the code for identification), and places the bread to trigger the weight sensor count. Real-time acquisition: cavity temperature is collected at 200ms intervals, and environmental temperature and humidity are read at 1 second intervals. The device ID, timestamp, and current stage information are uploaded to the cloud in JSON format.

[0044] Cloud algorithm processing: data preprocessing: cleaning abnormal values (such as temperature mutation ±20℃ is considered as sensor failure, triggering calibration), and normalizing environmental parameters (T_env mapped to [0,1]). Model calculation: ① Call environmental influence factor model: K_env = f(T_env, H_env) (XGBoost model trained based on 200,000 historical data); ② Calculate aging coefficient: α = 0.98^(usage days / 365) + 0.005×heating times (initial value α = 1, the more severe the aging, the smaller α); ③ Match the baking curve: according to the bread type, obtain the standard curve, and combine the current stage progress (such as the second minute in the constant baking stage) to generate the target temperature T_target. Generate control instructions: T_compensated = T_target × (1 + K_env ×(1-α)), convert to heating power P = k × T_compensated (k is the hardware calibration coefficient). After receiving the instructions at the device end, adjust the heating tube power (such as from 800W to 900W), and feedback the actual temperature in the next cycle. If the deviation is >5℃, trigger secondary compensation.

[0045] Aging compensation details include: initial calibration: record the initial heating efficiency (empty load temperature rise rate 5℃ / s) when the device is shipped, and record the actual temperature rise rate after each use, calculate the aging degree by Δrate = initial rate - current rate. Self-learning mechanism: when insufficient coloring occurs for 3 consecutive bakes, automatically trigger aging coefficient update (α increases by 0.05), without user intervention.

[0046] Through real-time temperature and humidity compensation, the temperature deviation during baking at -10°C is reduced from ±15°C in the traditional scheme to ±5°C, and the success rate of winter baking is increased from 70% to 95%. Based on the usage time and thermal efficiency decay model, the equipment used after 3 years can still maintain the initial baking quality, and the service life is extended by more than 20%. Combined with the number of bread (automatic increase of 10% power when 2 pieces), the stage progress (dynamic temperature adjustment in the coloring stage), realize "one piece one control, one stage one strategy", the burning rate is reduced from 15% to 3%. Through linkage with smart meters to realize heating during low valley period (electricity price reduced by 30%), linkage with kitchen air conditioner to maintain the best baking environment (humidity > 70% automatically start dehumidifier), build smart kitchen ecology.

[0047] Fusion of environmental parameters, equipment aging, and baking demand, build cross-domain correlation model, compared with traditional single-point feedback scheme, control dimension is expanded from 1D (temperature) to 5D (temperature, humidity, quantity, type, aging). Through massive device data continuous iteration algorithm (monthly model update), solve the problem of traditional single machine "the more you use, the less accurate" and realize "self-evolution" intelligence. Support APP remote setting baking preference, view device health status, abnormal situation (such as aging over limit) automatically push maintenance reminder, reduce the use threshold.

[0048] Unified cloud model reduces hardware calibration cost, and the performance consistency of different batches of equipment is improved by 40%. Smart meter linkage + power dynamic adjustment, compared with traditional toaster, energy saving 15%-20%, in line with energy efficiency standards.

[0049] The system breaks the device island through Internet of Things technology, uses data fusion and machine learning to build a dynamic compensation model, realizes the leap from "fixed program control" to "intelligent adaptive adjustment", effectively solves the core problems of traditional toaster such as poor environmental adaptation, no compensation for aging, and extensive control, and provides a replicable technical paradigm for fine control of intelligent home appliances.

[0050] In some embodiments, after generating the temperature compensation value at the current time, further comprising: through a dynamic time warping algorithm, deviation analysis is performed on the compensation cavity temperature data fed back by the toaster and the standard baking curve, the temperature compensation value is corrected in real time, and the compensation parameters of the algorithm model are updated.

[0051] After generating the temperature compensation value, a dynamic time warping algorithm (DTW) is introduced, time series alignment and deviation analysis are performed on the actual cavity temperature data fed back by the device and the standard baking curve, the compensation parameters are corrected in real time through a closed loop, and the algorithm model is self-optimized.

[0052] Time series alignment solves the slight deviation of temperature sampling frequency from the standard curve timeline in the actual baking process, ensuring the comparability of data in different stages. Deviation dynamic analysis calculates the cumulative distance (such as Euclidean distance) between the real-time temperature curve and the standard curve, identifying stages of insufficient or excessive compensation. Parameter iterative update adjusts the calculation parameters of environmental impact weight (K_env) and aging coefficient (a) according to the degree of deviation, forming a "compensation-feedback-correction" closed loop.

[0053] Data preprocessing: The device end collects the compensated cavity temperature sequence Treal(t) at 200ms intervals, and the cloud standard curve is an equal interval time sequence Tstd(t) (interval 1 second). Through the DTW algorithm, a time regularization matrix is constructed to find the optimal alignment path of the two sequences, eliminating the influence of sampling frequency difference.

[0054] Deviation calculation and classification: Calculate the point-by-point deviation ΔT(t)=Treal(t)−Tstd(t) of the aligned real-time curve and the standard curve, and set a threshold (such as ±3℃) to trigger correction. According to the baking stage (preheating, constant baking, coloring), the deviation is classified and counted, for example, if the continuous deviation in the coloring stage is >+5℃, it is determined as "overcompensation".

[0055] Parameter correction logic includes: If the deviation of the last 3 sampling points exceeds the threshold, adjust the model parameters according to the deviation direction: when the temperature is insufficient, increase the environmental impact weight Kenv+=0.05 or reduce the aging coefficient decay rate; when the temperature is too high, reduce Kenv−=0.03 or increase the aging coefficient decay. The corrected parameters are updated to the device end and cloud model, and the next baking directly applies the new parameters.

[0056] Through real-time curve alignment, the correction delay caused by time axis misalignment in traditional feedback mechanism is eliminated, and the temperature control error is reduced from ±5℃ to ±2℃. Without manual intervention, the system automatically iterates the compensation parameters according to the actual baking effect, solving the hardware differences of different batches of equipment and the non-linear aging problem in long-term use. The response speed to sudden environmental changes (such as opening the door to take bread causing temperature drop) is improved by 30%, and the abnormal stage is quickly located through DTW and the compensation strategy is dynamically adjusted.

[0057] In some embodiments, the intelligent temperature compensation algorithm model is constructed using machine learning algorithms, which can optimize the environmental parameter impact weight and the aging coefficient correction rule by continuously learning new baking data. The toaster terminal device, environmental perception linkage module, and computer device realize data interaction and linkage control through the Internet of Things platform.

[0058] The intelligent temperature compensation algorithm model is constructed based on machine learning (such as random forest, LSTM), and supports continuous learning of new baking data through the cloud to optimize the environmental parameter influence weight and the aging coefficient rule; the Internet of Things platform serves as a data hub, realizes three-way interaction of toasters, environmental sensors and the cloud, and constructs a collaborative architecture of "device collection-cloud training-edge execution".

[0059] The cloud server regularly (such as daily) summarizes the baking data of all network devices, updates the model parameters through online learning algorithm, and avoids the calculation overhead of retraining. The cloud server provides device management (registration, authentication), data storage (time series database InfluxDB), rule engine (trigger linkage instruction) and other services, and supports access through MQTT, HTTP and other protocols.

[0060] The machine learning model construction includes: offline training: the initial model is trained based on 200,000 sets of historical data, the input includes environmental temperature and humidity, device usage time, bread type, actual temperature deviation, and the output is the environmental influence weight Kenv and the aging coefficient a. After each baking is completed, the device uploads the complete temperature curve, compensation parameters and baking results (user score or automatic determination of success or failure), and the cloud adjusts the model weight through the gradient descent algorithm (learning rate 0.01). A new version of the model is generated every week and pushed to the device.

[0061] The Internet of Things platform interaction process: data upload: the device sends real-time data to the platform topic (such as "toaster / data / device123") through the MQTT protocol, and the platform stores it in the time series database after parsing. Instruction download: after the cloud model calculates the control instruction, it is sent to the device Topic through the platform rule engine, and the device subscribes to it in real time (QoS = 1 to ensure reliable transmission). Linkage control: when the platform detects that the environmental humidity is greater than 80%, it automatically sends a start instruction to the dehumidifier and triggers the toaster to increase the compensation power by 10%.

[0062] Through the accumulation of data from all network devices, the model's compensation ability for rare environments (such as high-altitude low pressure) gradually increases, solving the problem of poor generalization caused by insufficient data from a single machine. New devices can directly download the latest model when connected, eliminating the need for repeated calibration and improving production consistency by 50%. Old devices can be updated with the model to compensate for declining hardware performance and extend their service life. The Internet of Things platform supports scenario-based linkage (such as automatically adjusting the baking time according to the user's schedule and optimizing the compensation strategy based on weather forecasts), and builds an active intelligent service.

[0063] In some embodiments, the intelligent temperature compensation algorithm model formed by training historical baking data includes: multi-dimensional feature labeling of historically collected baking cavity temperature data, corresponding ambient temperature and humidity data, toaster usage time data, bread type and baking result data, and iterative training of the parameters of the ambient temperature influencing factor calculation unit, the cavity aging coefficient calculation unit and the bread category baking curve database using a gradient descent optimization algorithm to establish the intelligent temperature compensation algorithm model.

[0064] The training process of the intelligent temperature compensation algorithm model: historically collected multi-dimensional data (cavity temperature, ambient temperature and humidity, usage time, bread type, baking results) is feature-labeled, and a gradient descent optimization algorithm is used to iteratively train environmental influencing factors, aging coefficients, and baking curve parameters to form a quantifiable mathematical model.

[0065] Feature engineering extracts over 10 features, including ambient temperature difference (ΔT = T_env - 25°C, standard reference temperature), humidity influence coefficient (H_env / 50%, normalized), number of days used (t), and number of heating cycles (n). Labels are defined by categorizing baking results as "successful" (evenly browned), "inadequate" (underbaked), and "overbaked" (burned), which serve as label data for supervised learning.

[0066] Data labeling and preprocessing: 100,000 historical baking data points were annotated to extract peak temperatures, average deviations at each stage, and user feedback. Anomalous data (such as sudden temperature changes caused by sensor failures) was removed. Normalization involved mapping ambient temperature to the range [-1, 1] and performing a logarithmic transformation (ln(t+1)) on the duration to prevent differences in feature scales from impacting training results.

[0067] The model architecture and training use a multi-layer perceptron (MLP) as the underlying architecture, with a 10-dimensional feature input layer, two hidden layers (64 and 32 neurons), and an output layer containing the environmental impact weight Kenv and the aging coefficient α. The loss function is mean squared error (MSE) + cross entropy (classification baking result). The Adam optimizer (learning rate 0.001) is used for 200 training epochs, with a validation set accounting for 20%.

[0068] Iterative parameter optimization: Calculate the gradient ∇θ during each training cycle and update the model parameters θ←θ−η∇θ in the reverse direction of the gradient until the validation set loss converges (MSE < 0.05). Regularly (monthly) retrain the model with 10,000+ newly collected data to prevent overfitting (L2 regularization λ = 0.01).

[0069] Compared with traditional rule engines, the implicit correlation between the environment, equipment, and baking results is mined through machine learning (e.g., when the humidity is greater than 60% and the use is more than 2 years, the aging coefficient attenuation speed is accelerated by 20%), and the compensation strategy is more accurate. For unannotated new bread types (such as user-defined handmade bread), the model can give reasonable compensation suggestions through feature migration, avoiding the problem of "non-pre-set type cannot be compensated" in traditional solutions. Through feature importance analysis (such as SHAP value), it is clear that the influence of environmental temperature on compensation weight accounts for 40%, providing direction for subsequent model optimization.

[0070] In some embodiments, the real-time environmental parameter obtained by the environment perception linkage module is combined to calculate the environmental influence weight, including: according to the difference interval of real-time environmental temperature and standard reference temperature, the influence coefficient of environmental humidity on heat conduction efficiency, matching the corresponding weight adjustment factor through the preset environmental parameter influence rule library, the weight adjustment factor is associated with the historical baking temperature deviation data under the same environmental condition, forming a dynamic adjustable environmental temperature influence weight calculation mechanism of cavity heating efficiency, used for calculating the environmental influence weight.

[0071] The environmental influence weight calculation combines real-time environmental parameters (temperature, humidity) and historical deviation data through a hybrid mechanism of preset rule library + dynamic correction: rule library construction: divide 9 environmental scenes according to environmental temperature interval (such as <10℃, 10-25℃, >25℃) and humidity level (low, medium, high), and preset basic weight factor for each scene (such as low temperature and high humidity scene initial K_env=1.3). Dynamic correction: associate the historical baking temperature deviation mean under the same scene (such as the historical average under-baking of 5℃ in a certain scene, then the weight factor increases by 0.05), forming a double-layer mechanism of "rule base + data correction".

[0072] Environmental parameter classification: temperature interval: low temperature (<15℃), normal temperature (15-25℃), high temperature (>25℃); humidity level: dry (<40%RH), moderate (40%-60%RH), humid (>60%RH). Combined into 9 scenes, each scene stores the historical average deviation μΔT and the correction coefficient δ (initially 0).

[0073] Weight calculation process: Step 1: match the scene according to real-time T_env and H_env, get the basic weight Kbase (such as low temperature and high humidity scene Kbase=1.2). Step 2: query the historical deviation μΔT of the scene, calculate the dynamic correction term δ=μΔT / 10 (temperature deviation 10℃, weight adjustment 0.1). Step 3: the final weight K_env=K_base+δ, limit range [0.8, 1.5] to prevent extreme values. Rule library update: after each baking is completed, update the μΔT of the actual scene and deviation (sliding average, window size 100 times), continuously optimize the basic rules.

[0074] Compared with pure data-driven models, rule-based implementation achieves millisecond-level scene matching, is suitable for preheating stages with high real-time requirements, and avoids machine learning inference delays. By combining preset rules based on thermodynamic principles (humidity affects heat conduction efficiency, and low temperature requires higher heating power), the dependence on massive data is reduced, and the cold start performance is improved by 60%. For common kitchen environments (such as a winter kitchen at 10°C / 80%RH), through historical deviation correction, the compensation weight is dynamically adjusted from a fixed 1.2 to 1.35, directly solving the problem of decreased heating efficiency caused by low temperature and humidity.

[0075] In some embodiments, the cavity aging coefficient is calculated according to the toasting oven historical usage time, including: establishing an aging evaluation model based on the cumulative working time of the heating element, the loss degree of the cavity insulation layer, and the temperature sensor drift amount, comparing the slope difference between the standard heating curve of a new device and the actual heating curve of the current device, fitting the cavity thermal efficiency decay curve according to the usage time, and generating a cavity aging coefficient reflecting the degradation degree of the cavity heating performance.

[0076] The cavity aging coefficient calculation constructs a multi-dimensional aging evaluation model, which comprehensively considers the cumulative time of the heating element (the core aging factor), the loss of the insulation layer (the leakage rate of the cavity is detected by an infrared thermal imager), and the sensor drift (the offset amount is calibrated regularly). By comparing the slope difference between the standard curve of a new device and the actual curve of the current device, the thermal efficiency decay curve is fitted. The core formula is: α = 1−β×(t / t design life + 0.3γ+0.2δ); where β is the decay coefficient, γ is the insulation layer loss rate, and δ is the sensor drift degree.

[0077] Aging parameter acquisition: heating element time: the working time is accumulated every time the device is heated (accurate to seconds), and the design life is preset to 3000 hours. Insulation layer loss: at the first use of each year, the cavity shell temperature is detected by heating to 200°C under no load (leakage rate > 15% is determined as loss, γ = 1; otherwise, γ = 0). Sensor drift: automatically calibrated when starting, compare with the built-in standard heat source (100°C constant temperature block), record δ = 1 when the drift > ±2°C, otherwise δ = 0. Curve comparison and analysis: new device standard curve: under no load, the temperature rise rate is 5°C / s, and the temperature fluctuation in the stable stage is ±1°C. Current device actual curve: record the temperature rise rate vreal every time the preheating stage is recorded, and calculate the difference rate with the standard rate vstandard Δv = (vstandard−vreal) / vstandard. Aging coefficient fitting: through linear regression, α = 0.995 t / 30 (every 30 hours of use, α decays by 0.5%), combined with Δv dynamic adjustment (such as Δv > 20%, α decays by an additional 1%).

[0078] Self-diagnosis mechanism: when α < 0.7 (serious aging), the device pushes a replacement heating tube reminder, and the cloud automatically enhances compensation (increases power by 15%).

[0079] Break the defects of the traditional scheme "no aging compensation", convert the aging degree into a calculable α value through multi-dimensional parameters, realize the leap from "experience judgment" to "data quantification". After using the device for 3 years, through α dynamic compensation, the baking temperature deviation is controlled within ±4℃ from the traditional scheme's ±10℃, and the quality degradation rate is reduced by 70%. Through the aging coefficient warning, users can plan equipment maintenance in advance, reduce the probability of sudden failure, and reduce after-sales costs by 30%.

[0080] In some embodiments, the generation of the temperature compensation value at the current time based on the real-time data, the environmental influence weight, the cavity aging coefficient, and the standard baking curve of the corresponding bread category includes: according to the bread type parameter, the target temperature curve in the bread category baking curve database is called, and the target temperature curve is corrected according to the environmental temperature influence weight for the current baking stage progress; based on the cavity aging coefficient, the target temperature after correction is adjusted to form a real-time dynamic temperature compensation value.

[0081] Temperature compensation value generation integrates four elements: bread type (corresponding to standard curve), baking stage (different requirements for preheating / constant baking / coloring), environmental influence (K_env correction), and aging compensation (α adjustment), which realizes dynamic adaptation through hierarchical calculation. Core process: Curve matching: according to the bread type, the corresponding temperature-time curve in the database is called (such as baguette which needs high temperature and short time baking at 220℃).

[0082] Environmental correction: adjust the target temperature of each stage according to the current environmental weight K_env (increase the target value in low temperature environment).

[0083] Aging compensation: based on the α value, reduce the temperature loss caused by the heat efficiency decay (increase the compensation power when the aging is serious).

[0084] Standard curve structure: Each curve contains 3 stage parameters: preheating segment (target temperature T1, time t1), constant baking segment (T2, t2), and coloring segment (T3, t3), for example, white bread curve: T1=150℃ (t1=1min), T2=180℃ (t2=2min), T3=200℃ (t2=1min).

[0085] The hierarchical calculation logic includes: Step 1: Stage positioning: determine the stage according to the current heating time (such as entering the coloring stage at the 3rd minute), and obtain the standard target temperature Tstd of the stage. Step 2: Environmental correction: Tenv_corr = Tstd x Kenv (when K_env > 1, increase the target temperature to cope with low temperature environment). Step 3: Aging compensation: Tfinal = Tenv_corr + (1 - a) x 10℃ (the smaller a is, the greater the compensation is, and the empirical coefficient 10℃ is fitted through historical data). Step 4: Quantity correction: if 2 pieces of bread are detected, Tfinal += 5℃ (increase the heat load compensation).

[0086] Boundary condition processing: temperature upper limit 250℃: when exceeded, forced truncation, converted to extended heating time compensation; power upper limit: adjusted according to real-time power limit of smart meter to avoid overload (such as when the household circuit limit is 2000W, the power does not exceed 1800W).

[0087] From the "one-size-fits-all" fixed program to the "stage + environment + aging + quantity" four-dimensional dynamic compensation, for example, when 4 pieces of bread are placed at the same time, the system automatically identifies and increases the power by 20% in the constant baking stage to avoid underbaking of the middle bread. Support unlimited addition of bread types, just enter their standard curves in the database, and they can be automatically adapted through the environmental and aging compensation mechanism, shortening the new product development cycle by 50%. Through dynamic power adjustment, on the premise of meeting the baking quality, avoid the traditional scheme of excessive heating to cope with extreme environments, energy saving effect is improved by 18%, and the charring rate is reduced to below 2%.

[0088] In some embodiments, by introducing user individual preference parameters in the temperature compensation model, by analyzing user historical baking score data, a "taste preference-compensation parameter" mapping relationship is established to realize differentiated compensation of crispness and coloring depth for different users.

[0089] Preference feature modeling: convert user scores (1-5) into crispness factor (Crispness Factor, CF) and coloring factor (Color Factor, CF) as dynamic inputs of the compensation model. Real-time preference adaptation: according to the "soft / crisp" taste label selected by the current user, adjust the heating rate and peak temperature compensation weight of the standard baking curve in real time.

[0090] Preference data collection and training: Users rate the toasting result through the APP (with "too soft / just right / too crisp" labels), after 5 or more ratings, use K-means algorithm to classify users into "soft texture type", "balanced type", "crisp texture type". Build preference neural network model: input is historical compensation parameters (K_env, a), bread type, rating label, output is personalized compensation offset (ΔT_crisp, ΔT_color).

[0091] Real-time compensation process: Users select texture preference when starting toasting (default balanced type), the device retrieves the corresponding compensation offset. For example, "crisp texture type" user: automatically increase the target temperature by 5-10°C in the browning stage, and increase the compensation parameter K_env by 0.1 (accelerate water evaporation). After each rating, the model updates the offset through small batch gradient descent (learning rate 0.005) to adapt to individual subtle preference changes.

[0092] Edge storage mechanism: Store the user's last 10 preference adjustment records locally on the device, still provide personalized compensation based on historical data when offline, and synchronize to the cloud model when online.

[0093] Solve the problem of "different user needs on the same device", for example, children users prefer soft bread (reduce peak temperature compensation parameter), elderly users prefer crisp texture (increase browning compensation), user satisfaction is improved by 40%. No need for users to manually set parameters, automatically learn through behavior data, especially suitable for baking beginners and high-end user segmentation scenarios. Based on user preference data, customized bread raw material purchase suggestions can be pushed (such as crisp texture users recommended high starch bread flour), forming a "device-user-supply chain" closed loop.

[0094] In some embodiments, by building a digital twin model of the toaster cavity, driving the temperature field simulation of the virtual cavity through real-time sensor data, predicting the toasting result under the current compensation strategy, and correcting the compensation parameters in advance.

[0095] Physical field modeling: Based on finite element analysis (FEA), establish a three-dimensional simulation model of cavity heat conduction and air convection, input heating tube power, environmental parameters, and bread placement position. Virtual-real mapping calibration: Calibrate the virtual model's material thermal conductivity, radiation coefficient, and other parameters through actual toasting data (every 10 times) to ensure simulation error <2%.

[0096] Digital twin model construction: Use COMSOL Multiphysics to establish a cavity geometry model, input material properties after meshing (stainless steel shell thermal conductivity 16W / mK, ceramic inner shell 5W / mK). Define boundary conditions: heating tube heat source distribution, bread heat capacity (1.8kJ / kg*K), cavity and environment convective heat transfer coefficient (5W / m2 K).

[0097] Pre-compensation simulation process: Before the start of baking, the user places the bread, and the device detects the number and position of the bread through the gravity sensor, and inputs the current environmental parameters (T=20℃, H=50%) and initial compensation parameters (K_env=1.1, α=0.95) into the digital twin model. The simulation predicts the cavity temperature distribution after 3 minutes.

[0098] If the simulation shows that the center temperature of the bread does not reach the target (deviation > 5℃), automatically adjust K_env to 1.2, re-simulate until the error is <2℃, and then start heating.

[0099] Online calibration mechanism: After each baking is completed, compare the actual temperature curve with the simulation curve, and use the Bayesian optimization algorithm to update the unknown parameters such as the convection heat transfer coefficient in the model, continuously improving the simulation accuracy.

[0100] From "feedback correction" to "feedforward optimization", the first-time baking success rate is increased from 75% to 92%, especially solving the compensation problem of complex scenarios (multi-layer bread rack, different placement positions). Through pre-simulation, invalid heating is avoided, and 15% of electric energy is saved on average for each baking, while the preheating time is shortened by 1-2 minutes. The digital twin model can be used for new model design verification, optimizing the cavity structure before physical prototype manufacturing, and shortening the development cycle by 30%.

[0101] In some embodiments, by meeting the user's privacy protection needs, federated learning (Federated Learning) technology is used to realize the collaborative optimization of compensation models for all network devices without sharing the original baking data.

[0102] Hierarchical training architecture: The device side trains the compensation model parameters locally, and the cloud side aggregates the model gradients (non-original data), protecting the privacy of the user's baking habits. Differentiated weight aggregation: According to the differences in the use environment of the device (such as the humidity difference between the north and south), different weights are given to the model updates of devices in different regions, improving the regional compensation relevance.

[0103] Federal learning process design: local training: after each baking, the device calculates the model gradient (AK_env, Delta a) using local data (environment parameters, compensation parameters, baking results), and uploads it to the cloud after encryption. Federal aggregation: the cloud groups by region (such as North China and South China), and performs weighted averaging (weight = historical data quality score of the device) on the gradients of devices in the same region to generate a region-specific compensation model. Model distribution: the device downloads the updated model for the corresponding region while retaining 10% of the local personalized parameters (such as user preferences) that do not participate in aggregation. Privacy protection mechanism: add differential privacy noise (Laplace mechanism, epsilon = 0.5) to the gradient data to prevent user data from being reverse-engineered. Use homomorphic encryption technology to ensure that the original gradient content cannot be decrypted when aggregated in the cloud. Dynamic grouping strategy: the device automatically joins the corresponding regional group based on the IP address when it first connects to the network, and is re-grouped every quarter (based on environmental parameter changes).

[0104] Comply with data protection regulations, user data does not leave the device, solve the privacy leakage risk of traditional cloud training. The model in the dry northern region is automatically enhanced for low-temperature compensation (the average K_env is 0.08 higher than in the south), solving the adaptability problem of the "one-size-fits-all" model in different climate zones, and improving the accuracy of regional compensation by 25%. Even with a small amount of data from a single device (such as a commercial toaster with low usage frequency), the federated aggregation can obtain the experience of the entire network, avoiding "cold start" compensation failure.

[0105] In some embodiments, by introducing a deep reinforcement learning (DRL) architecture, the temperature compensation process is modeled as a sequential decision problem, and the optimal compensation strategy is autonomously explored through a "state-action-reward" loop without a pre-set rule base. State space definition: contains 15-dimensional state variables including current cavity temperature, environmental parameters, baking stage, remaining time, bread type, etc. Action space design: discrete actions include adjusting heating power (10 levels), compensation temperature offset (±15°C), maintaining the current strategy, and each decision interval is 2 seconds. Reward function design: the deviation square sum of the actual temperature and the standard curve is taken as the negative reward, and the "early completion of baking" is taken as the positive reward, guiding the algorithm to balance accuracy and efficiency.

[0106] DRL model construction: use a deep Q network (DQN) architecture, input layer 15-dimensional state, hidden layer 2 layers (each with 128 neurons), output layer 10-dimensional action value. Experience replay pool stores historical state-action-reward data, updates the target network every 50 decisions to prevent unstable training.

[0107] Online learning process: When the device is running in "learning mode", DRL exploration is turned on (epsilon-greedy strategy, initial epsilon=0.9, decay to 0.1 with training), real-time attempts are made to different compensation actions. After each baking is completed, the cumulative reward is calculated according to the final score (success + 100 points, burning - 200 points), and the Q network parameters are updated (learning rate 0.001). In normal user mode, switch to "inference mode" and use the trained strategy to directly output the optimal compensation action.

[0108] Safety constraint design: Action space limitation: power does not exceed 120% of the rated value, temperature compensation does not exceed ±20℃, to avoid hardware damage.

[0109] Break through the limitations of traditional "preset rules + data correction", automatically discover compensation strategies that human experts have not noticed (such as high-altitude area stage compensation strategy), and improve the compensation effect in complex scenarios by 30%. The response speed to sudden interference (such as frequent opening of the kitchen door causing environmental temperature fluctuations) is shortened from 10 seconds in the traditional scheme to 3 seconds, and a "disturbance-compensation" conditioned reflex is formed through reinforcement learning. With the accumulation of learning data from all network devices, the DRL strategy library is constantly enriched, and it can cope with the baking needs of new types of bread materials (such as plant-based bread) in the future without the need for manual reprogramming.

[0110] In some embodiments, by integrating a miniature camera inside the toaster, real-time monitoring of the bread baking state (coloring degree, expansion rate) is realized through computer vision technology, and the compensation parameters are dynamically adjusted to realize a "visual feedback-intelligent compensation" closed loop.

[0111] Image feature extraction: Use a lightweight CNN model (such as MobileNetV3) to identify the bread surface spot area ratio and color RGB value changes in real time. State mapping algorithm: Establish a "visual feature-temperature compensation" mapping relationship, for example, when the spot ratio exceeds 30%, automatically reduce the compensation value by 5℃.

[0112] A 720P miniature camera is installed on the top of the toaster cavity, the lens is treated with oil-proofing, and an image is captured every 10 seconds. The edge computing module runs a real-time image analysis model to calculate the current coloring index (a * value in CIELAB color space, reflecting the degree of redness) and the spot ratio. Dynamic compensation logic: preset the coloring target interval (such as a * =25-35 is the best), when a * < 25 and the spot ratio < 10%, it is determined that the "coloring is insufficient", and K_env=0.05 is increased; when a *When the >35 or the focal spot >20% is determined as "over-coloring", the current stage target temperature is reduced by 10℃, and the subsequent stage time is extended. The baking stage adjustment strategy is combined: the preheating stage mainly refers to the expansion rate (the volume change is calculated by contour detection), and the coloring stage focuses on the color features. Anti-interference processing: background subtraction algorithm is used to remove the interference of cavity reflection, and time series smoothing filtering is used to filter accidental noise (such as bread crumb reflection).

[0113] Upgrading from "parameter-based compensation" to "result-based compensation", directly adjusting to the actual state of the bread, solving the compensation deviation caused by the difference of raw materials of different brands of bread, and reducing the burning rate from 8% to 1.5%. Users do not need to select bread types, and the device automatically identifies categories such as toast, baguette, and brioche (recognition accuracy 92%) and matches the corresponding visual compensation strategy, truly realizing "just put and bake". Camera data can be used to detect heating tube abnormalities (such as one-sided non-emission leading to uneven coloring), and combined with image recognition, the device realizes visual diagnosis of hardware faults, and the maintenance efficiency is improved by 50%.

[0114] Please refer to Figure 2 , Figure 2 is a schematic flowchart of an Internet of Things linked toasty oven intelligent temperature compensation method provided in an embodiment of the present application. The execution device of the method is a computer device of the Internet of Things linked toasty oven intelligent temperature compensation system provided in any embodiment of the present application.

[0115] As shown in Figure 2 , the provided method includes steps S101 to S102. The computer device can be a handheld terminal, a notebook computer, a wearable device, or a robot, etc. The steps S101 to S102 and the corresponding embodiments thereof are used to implement.

[0116] Step S101. An intelligent temperature compensation algorithm model is formed by training historical baking data, the intelligent temperature compensation algorithm model at least includes an environmental temperature influence factor calculation unit, a cavity aging coefficient calculation unit, and a bread category baking curve database, the environmental temperature influence factor calculation unit is used to calculate the environmental influence weight combined with the real-time environmental parameters obtained by the environmental perception linkage module, and the cavity aging coefficient calculation unit is used to calculate the cavity aging coefficient according to the historical use time of the toasty oven; Step S102. Based on the real-time data of the toasty oven, the environmental influence weight, the cavity aging coefficient, and the standard baking curve of the corresponding bread category, a temperature compensation value at the current time is generated.

[0117] In some embodiments, after the temperature compensation value at the current time is generated, it further includes: deviation analysis of the compensation cavity temperature data fed back by the toasty oven and the standard baking curve by dynamic time warping algorithm, real-time correction of the temperature compensation value and update of the compensation parameters of the algorithm model.

[0118] In some embodiments, the intelligent temperature compensation algorithm model is constructed by a machine learning algorithm, which can optimize the environmental parameter influence weight and the aging coefficient correction rule through continuous learning of new baking data. The toaster terminal device, the environment perception linkage module, and the computer device realize data interaction and linkage control through an Internet of Things platform.

[0119] In some embodiments, the intelligent temperature compensation algorithm model formed by training historical baking data includes: multi-dimensional feature labeling of the historical baking cavity temperature data, corresponding environmental temperature and humidity data, toaster usage time data, bread type, and baking result data; iterative training of the parameters of the environmental temperature influence factor calculation unit, the cavity aging coefficient calculation unit, and the bread category baking curve database using a gradient descent optimization algorithm; and establishing the intelligent temperature compensation algorithm model.

[0120] In some embodiments, calculating the environmental influence weight based on the real-time environmental parameters obtained by the environment perception linkage module includes: matching corresponding weight adjustment factors through a preset environmental parameter influence rule base based on the difference interval between the real-time environmental temperature and the standard reference temperature and the influence coefficient of environmental humidity on heat conduction efficiency; the weight adjustment factors are associated with historical baking temperature deviation data under the same environmental conditions, forming a dynamic adjustable environmental temperature influence weight calculation mechanism for the cavity heating efficiency, which is used to calculate the environmental influence weight.

[0121] In some embodiments, calculating the cavity aging coefficient based on the historical usage time of the toaster includes: establishing an aging evaluation model based on the cumulative working time of the heating element, the loss degree of the cavity insulation layer, and the temperature sensor drift amount; fitting the cavity thermal efficiency decay curve by comparing the slope difference between the standard heating curve of a new device and the actual heating curve of the current device, and generating a cavity aging coefficient reflecting the degradation degree of the cavity heating performance.

[0122] In some embodiments, generating a temperature compensation value at the current time based on the real-time data, the environmental influence weight, the cavity aging coefficient, and the standard baking curve of the corresponding bread category includes: matching and calling the target temperature curve in the bread category baking curve database according to the bread type parameter; and performing environmental temperature correction on the target temperature curve based on the environmental influence weight for the current baking stage progress; and performing thermal efficiency compensation adjustment on the corrected target temperature based on the cavity aging coefficient to form a real-time dynamic temperature compensation value.

[0123] It should be noted that, for the convenience and brevity of description, the specific working process of the above-described IoT linkage toaster intelligent temperature compensation method and each step can refer to the corresponding process in the IoT linkage toaster intelligent temperature compensation system embodiments described in the above embodiments, and will not be repeated here.

[0124] Please refer to Figure 3 , Figure 3 is a structural schematic block diagram of a computer device provided by the embodiment of the present application. The computer device comprises a processor, a memory and a network interface connected through a device bus, wherein the memory can comprise a storage medium and an internal memory.

[0125] The storage medium can store an operating device and a computer program. The computer program comprises program instructions which, when executed, can cause the processor to execute any one of the embodiments of the IoT linkage toaster intelligent temperature compensation method.

[0126] The processor is used to provide computing and control capabilities to support the operation of the entire computer device.

[0127] The internal memory provides an environment for the running of the computer program in the non-volatile storage medium, which, when executed by the processor, can cause the processor to execute any one of the IoT linkage toaster intelligent temperature compensation system methods.

[0128] The network interface is used for network communication, such as sending assigned tasks, etc. Those skilled in the art can understand that Figure 3 The structure shown in the above

[0129] It should be understood that the processor can be a central processing unit (CPU), and the processor can also be other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), field programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among them, the general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc.

[0130] In one embodiment, the processor is configured to run a computer program stored in the memory to implement the following steps: The intelligent temperature compensation algorithm model formed by training historical baking data comprises at least an environmental temperature influence factor calculation unit, a cavity aging coefficient calculation unit, and a bread category baking curve database, the environmental temperature influence factor calculation unit is configured to calculate an environmental influence weight based on real-time environmental parameters obtained by the environmental perception linkage module, and the cavity aging coefficient calculation unit is configured to calculate a cavity aging coefficient based on a historical use time length of the toaster. Based on the real-time data of the toaster, the environmental influence weight, the cavity aging coefficient, and the standard baking curve of the corresponding bread category, a temperature compensation value at the current time is generated.

[0131] In some embodiments, after the temperature compensation value at the current time is generated, the method further comprises: performing deviation analysis on the compensated cavity temperature data fed back by the toaster and the standard baking curve by a dynamic time warping algorithm, and updating the compensation parameters of the algorithm model by real-time correcting the temperature compensation value.

[0132] In some embodiments, the intelligent temperature compensation algorithm model is constructed by using a machine learning algorithm, and can optimize the environmental parameter influence weight and the aging coefficient correction rule by continuously learning new baking data. The toaster terminal device, the environmental perception linkage module, and the computer device realize data interaction and linkage control through an Internet of Things platform.

[0133] In some embodiments, the intelligent temperature compensation algorithm model formed by training historical baking data comprises: multi-dimensional feature labeling of historical baking cavity temperature data, corresponding environmental temperature and humidity data, toaster use time length data, bread type, and baking result data, iterative training of parameters of the environmental temperature influence factor calculation unit, the cavity aging coefficient calculation unit, and the bread category baking curve database by using a gradient descent optimization algorithm, and establishment of the intelligent temperature compensation algorithm model.

[0134] In some embodiments, calculating the environmental influence weight based on the real-time environmental parameters obtained by the environmental perception linkage module comprises: matching corresponding weight adjustment factors through a preset environmental parameter influence rule library based on a difference interval between a real-time environmental temperature and a standard reference temperature and an influence coefficient of environmental humidity on heat conduction efficiency, the weight adjustment factors being associated with historical baking temperature deviation data under the same environmental conditions, forming a dynamic adjustable environmental temperature influence weight calculation mechanism for cavity heating efficiency, and being used for calculating the environmental influence weight.

[0135] In some embodiments, the calculating the cavity aging coefficient according to the toaster historical usage time length comprises: establishing an aging evaluation model based on the cumulative working time length of the heating element, the loss degree of the cavity heat preservation layer, and the temperature sensor drift amount; fitting a cavity heat efficiency attenuation curve by comparing the slope difference between the new device standard heating curve and the current device actual heating curve, and combining the usage time length; and generating a cavity aging coefficient reflecting the degradation degree of the cavity heating performance.

[0136] In some embodiments, the generating the temperature compensation value at the current time based on the real-time data, the environmental influence weight, the cavity aging coefficient, and the standard toasting curve of the corresponding bread type comprises: calling a target temperature curve in the bread type toasting curve database according to the bread type parameters; performing environmental temperature correction on the target temperature curve according to the current toasting stage progress and the environmental influence weight; and performing heat efficiency compensation adjustment on the corrected target temperature based on the cavity aging coefficient to form a real-time dynamic temperature compensation value.

[0137] It should be noted that, for the convenience and brevity of description, the specific working process of the processor described above can refer to the corresponding process in the method embodiments described in the above embodiments, which will not be described here.

[0138] In the embodiments of the present application, a computer readable storage medium is also provided, which stores a computer program. The computer program comprises program instructions. The processor executes the program instructions to implement the steps of the Internet of Things linked toaster intelligent temperature compensation method provided in the above embodiments of the present application.

[0139] The computer readable storage medium can be an internal storage unit of the computer device, such as a hard disk or a memory of the computer device. The computer readable storage medium can also be an external storage device of the computer device, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc.

[0140] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of various equivalent modifications or replacements within the technical scope disclosed in the present application, and these modifications or replacements should be covered in the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. An Internet of Things-linked toaster intelligent temperature compensation system, characterized in that: include: A data acquisition module is used to obtain real-time data of the toaster; the real-time data includes the temperature of the baking cavity, the number of breads placed, the type of bread parameters and the progress of the current baking stage; An Internet of Things communication module, configured to transmit the real-time data and device identification to a computer device and receive control instructions issued by the computer device; A heating control module, configured to adjust the power output or heating time of the heating element of the toaster according to the control instruction; An environmental perception and linkage module is configured to obtain real-time ambient temperature and humidity data of the toaster's environment and support data interaction with surrounding IoT devices such as smart meters and ambient temperature and humidity sensors; A computer device having a built-in intelligent temperature compensation algorithm model trained using historical baking data, the intelligent temperature compensation algorithm model comprising at least an ambient temperature impact factor calculation unit, a cavity aging coefficient calculation unit, and a bread category baking curve database, the ambient temperature impact factor calculation unit being configured to calculate an environmental impact weight in combination with real-time environmental parameters acquired by the environmental perception linkage module, and the cavity aging coefficient calculation unit being configured to calculate a cavity aging coefficient based on the historical usage time of the toaster; Based on the real-time data, the environmental impact weight, the cavity aging coefficient and the standard baking curve of the corresponding bread category, a temperature compensation value at the current moment is generated.

2. The system according to claim 1, wherein: After generating the temperature compensation value at the current moment, the method further includes: The dynamic time warping algorithm is used to analyze the deviation between the compensated cavity temperature data fed back by the toaster and the standard baking curve, and the temperature compensation value is corrected in real time and the compensation parameters of the algorithm model are updated.

3. The system according to claim 1, wherein: The intelligent temperature compensation algorithm model is constructed using a machine learning algorithm, and can optimize the environmental parameter influence weights and aging coefficient correction rules by continuously learning new baking data. The toaster terminal device, environmental perception linkage module and computer equipment realize data interaction and linkage control through the Internet of Things platform.

4. The system according to claim 1, wherein: The intelligent temperature compensation algorithm model formed by training historical baking data includes: The historically collected baking cavity temperature data, corresponding ambient temperature and humidity data, toaster usage time data, bread type and baking result data are annotated with multi-dimensional features, and the gradient descent optimization algorithm is used to iteratively train the parameters of the ambient temperature influencing factor calculation unit, the cavity aging coefficient calculation unit and the bread category baking curve database to establish the intelligent temperature compensation algorithm model.

5. The system according to claim 1, wherein: The calculating of the environmental impact weight based on the real-time environmental parameters obtained by the environmental perception linkage module includes: According to the difference range between the real-time ambient temperature and the standard reference temperature, and the influence coefficient of ambient humidity on the heat conduction efficiency, the corresponding weight adjustment factor is matched through the preset environmental parameter influence rule library. The weight adjustment factor is associated with the historical baking temperature deviation data under the same environmental conditions to form a dynamically adjustable ambient temperature influence weight calculation mechanism for the cavity heating efficiency, which is used to calculate the environmental impact weight.

6. The system according to claim 1, wherein: The calculation of the cavity aging coefficient based on the historical usage time of the toaster includes: An aging assessment model is established based on the cumulative working hours of the heating element, the degree of loss of the cavity insulation layer and the drift of the temperature sensor. By comparing the slope difference between the standard heating curve of the new equipment and the actual heating curve of the current equipment, and combining the usage time to fit the cavity thermal efficiency attenuation curve, a cavity aging coefficient reflecting the degree of degradation of the cavity heating performance is generated.

7. The system according to claim 1, wherein: The method of generating a temperature compensation value at a current moment based on the real-time data, the environmental impact weight, the cavity aging coefficient, and the standard baking curve of the corresponding bread category includes: The target temperature curve in the bread category baking curve database is called based on the bread type parameters. Based on the current baking stage progress, the target temperature curve is corrected for the ambient temperature using the environmental impact weight. The corrected target temperature is adjusted for thermal efficiency compensation based on the cavity aging coefficient to form a real-time dynamic temperature compensation value.

8. An Internet of Things-linked toaster intelligent temperature compensation method, characterized in that: A computer device applied to an Internet of Things-linked toaster intelligent temperature compensation system according to any one of claims 1 to 7, wherein the method comprises: An intelligent temperature compensation algorithm model trained using historical baking data includes at least an ambient temperature impact factor calculation unit, a cavity aging coefficient calculation unit, and a bread baking curve database. The ambient temperature impact factor calculation unit is used to calculate the environmental impact weight based on real-time environmental parameters acquired by the environmental perception linkage module. The cavity aging coefficient calculation unit is used to calculate the cavity aging coefficient based on the toaster's historical usage time. The current temperature compensation value is generated based on the toaster's real-time data, environmental impact weight, cavity aging coefficient, and standard baking curve for the corresponding bread category.

9. A computer device, characterized in that: The computer device includes a memory and a processor; The memory is used to store computer programs; The processor is configured to execute the computer program and implement the method according to claim 8 when executing the computer program.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, which, when executed by a processor, causes the processor to implement the method according to claim 8 .

Images