Chef machine intelligent control method and system based on Internet of Things

Abstract

The invention discloses a chef machine intelligent control method and system based on the Internet of Things, and relates to the technical field of chef machines, and the method comprises the steps: building a two-dimensional thermophysical property distribution diagram of a pot bottom based on a cook machine pot structure and manufacturing data, loading a heat conduction finite element simulation model, simulating the thermal diffusion behavior of a pot body in an embedded edge calculation module, and calculating the thermal diffusion behavior of the pot body; and spatial registration and difference operation are performed by combining a temperature distribution diagram acquired by the thermal infrared imager in real time, so that online self-calibration of thermal field simulation is realized. Then dividing the pan bottom into asymmetric thermosensitive cells through a region growing algorithm and a weighted Voronoi strategy, and extracting multi-dimensional thermal sensing data including a temperature value, thermal capacity density, image disturbance characteristics and the like by taking the highest temperature point in each cell as a thermal sensing acquisition point; the mechanism not only improves the modeling precision of the pot bottom heterogeneity thermal response, but also effectively improves the hot spot identification and interference elimination capability.

Description

Technical Field

[0001] The present invention relates to the technical field of cooking machines, and specifically to an intelligent control method and system for a cooking machine based on the Internet of Things. Background Art

[0002] In an industrial automation system, an industrial switch is a core network device responsible for connecting and managing various industrial devices such as PLCs, sensors, and robots, ensuring the real-time and reliable transmission of data. In this specific field, data security issues are particularly important. The industrial switch not only has to process high-frequency data streams but also must ensure the security and stability of data transmission. With the rapid development of the industrial Internet and intelligent manufacturing, industrial switches need to cope with more network security threats such as packet loss, network latency, malicious traffic, and network attacks.

[0003] At the present stage, existing cooking machines lack a real-time perception mechanism based on infrared thermal imaging and are unable to construct a complete, dynamic, and reliable thermal field model for the bottom of the pot. Without considering the heat conduction distribution of the pot body material and the characteristics of local thermal inertia, relying solely on single-point temperature feedback will lead to control lag or misjudgment. When there are strong disturbances in the thermal image such as metal reflection or liquid surface steam, this image information is misidentified as abnormal temperature hotspots, and the system may incorrectly execute a power reduction operation, resulting in insufficient temperature in a local area, further causing undercooked ingredients or power waste. At the same time, it is unable to timely identify the sudden change area of temperature difference and real hotspots, easily leading to the accumulation of hotspots and causing phenomena such as bottom of the pot pasting, sticking, or over-frying. Therefore, there is an urgent need for a thermal control system that can integrate heat conduction modeling, infrared thermal data acquisition, disturbance discrimination, and intelligent control strategy linkage to solve the limitations of traditional temperature control solutions in terms of accuracy, energy efficiency, and intelligent judgment dimensions. Summary of the Invention

[0004] In view of the deficiencies of the prior art, the present invention provides an intelligent control method and system for a cooking machine based on the Internet of Things, which solves the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention is realized through the following technical solutions: including the following steps:

[0006] S1. Divide the bottom of the cooking machine pot into asymmetric thermosensitive cells, set the hottest point in the asymmetric thermosensitive cell as the collection point, and record the thermal perception data of each collection point based on the collection point;

[0007] S2. Construct a cloud server, set up the Internet of Things to transmit the collected thermal perception data to the cloud server, and preprocess the thermal perception data in the cloud server to obtain a standard thermal perception data set;

[0008] S3. Based on the standard thermal perception dataset, calculate and output the locally thermally anomalous corrected temperature difference index ΔT, traverse the locally thermally anomalous corrected temperature difference index ΔT between all pairwise acquisition points, and then summarize to obtain the corrected temperature difference index set FΔT;

[0009] S4. Set the multi-dimensional temperature control strategy threshold, based on the multi-dimensional temperature control strategy threshold, divide all the locally thermally anomalous corrected temperature difference indexes ΔT in the corrected temperature difference index set FΔT into regions, divide different asymmetric thermosensitive cells at the bottom of the cook machine pot into different response regions, and generate the level label L;

[0010] S5. Based on the level label L, calculate and output the control power value Q of the heating module at each acquisition point respectively, and then based on the control power value Q and combined with the standard thermal perception dataset, calculate and output the comprehensive evaluation value Rk of the intelligent thermal control effect, and set the effect threshold Rth to compare and evaluate with the comprehensive evaluation value Rk of the intelligent thermal control effect to analyze the temperature control situation.

[0011] Preferably, the S1 includes S11 and S12;

[0012] S11. According to the design drawings and manufacturing process data of the cook machine pot, obtain the material composition information and thickness parameters of each area at the bottom of the pot. Through the preset material database, map each area to the corresponding thermal conductivity k, specific heat capacity cp, and heat capacity density value ρ, calculate and obtain the thermal diffusivity a = k / ρ·cp, and construct a thermal physical property distribution map in the coordinate system of the two-dimensional cook machine pot bottom;

[0013] In the embedded edge computing module inside the cook machine, load the two-dimensional heat conduction finite element simulation model, set the bottom boundary conditions and heating area heat source input based on the thermal physical property distribution map, simulate the transient heat diffusion path of the cook machine pot bottom under the unit heating power, and generate a simulation thermal field distribution map;

[0014] When the temperature distribution map of the cook machine pot bottom is collected in real time by the infrared thermal imager, perform spatial registration and difference operation on the temperature distribution map of the cook machine pot bottom and the simulation thermal field distribution map to obtain a thermal field difference map, and perform online self-calibration on the thermal simulation model;

[0015] After the simulation online self-calibration, by introducing the region growing algorithm and weighted Voronoi partitioning strategy, subdivide the grid in the high thermal gradient region, merge and simplify the region with stable residual heat, and divide the cook machine pot bottom into multiple asymmetric thermosensitive cells;

[0016] S12. Based on the asymmetric thermosensitive cells, automatically identify the highest temperature point in each asymmetric thermosensitive cell, set the highest temperature point as the acquisition point of the asymmetric thermosensitive cell, then summarize all the acquisition points to obtain the thermal sensing point set P, and at the same time, each acquisition point records the thermal perception data in different asymmetric thermosensitive cells in real time;

[0017] The thermal perception data includes the temperature value T of the acquisition point, the heat capacity density value ρ, the local variance Varblock, the edge strength Edge, and the local information entropy Entropy.

[0018] Preferably, S2 includes S21 and S22;

[0019] S21. Build a cloud server for the chef machine through the Internet of Things, and use the loT Internet of Things technology to transmit the thermal perception data to the cloud server;

[0020] S22. In the cloud server, preprocess the thermal perception data to obtain a standard thermal perception data set, where the standard thermal perception data set includes the temperature value T of the acquisition point, the heat capacity density value ρ, the heating rate ψ, and the infrared image reflection disturbance factor φ;

[0021] The preprocessing includes data standardization processing and data feature extraction;

[0022] The data standardization processing uses Z-score standardization to perform normalization standardization processing on the thermal perception data, eliminating the dimensional influence between all parameters in the thermal perception data;

[0023] The data feature extraction obtains the heating rate ψ and the infrared image reflection disturbance factor φ by performing feature extraction based on the thermal perception data.

[0024] Preferably, S3 includes S31 and S32;

[0025] S31. Based on the obtained standard thermal perception data set, calculate and output the local thermal anomaly correction temperature difference index △T between two acquisition points, and analyze the degree of thermal anomaly between each pair of acquisition points;

[0026] The local thermal anomaly correction temperature difference index △T is calculated and output through the following algorithm formula;

[0027]

[0028] In the formula, △T ij represents the local thermal anomaly correction temperature difference index between the i-th acquisition point and the j-th acquisition point, φ i represents the infrared image reflection disturbance factor of the i-th acquisition point, φ j represents the infrared image reflection disturbance factor of the j-th point, ρ i represents the heat capacity density value of the i-th acquisition point, ρ j represents the heat capacity density value of the j-th acquisition point.

[0029] Preferably, in S32, traverse the standard thermal perception data set of the acquisition points in all asymmetric thermal-sensitive cells, calculate and output the local thermal anomaly correction temperature difference index ΔT of the acquisition points in all thermal sense point sets, and generate and summarize all the local thermal anomaly correction temperature difference indexes ΔT output by the acquisition points in all thermal sense point sets to obtain the correction temperature difference index set FΔT;

[0030] The expression of the correction temperature difference index set FΔT is: Among them, It means that all acquisition points i and acquisition point j are a pair of points in the thermal sense point set P.

[0031] Preferably, the S4 includes S41 and S42;

[0032] S41, the multi-dimensional temperature control strategy threshold includes a temperature upper limit threshold F1, a heating rate threshold F2, a medium temperature difference threshold F3, and a heating stability threshold F4;

[0033] The temperature upper limit threshold F1 is set by the maximum allowable temperature difference between the center and the edge of the cook machine pot;

[0034] The heating rate threshold F2 is set by the upper limit value of the heating rate ψ of the cook machine at different cooking modes;

[0035] The medium temperature difference threshold F3 is set by 50% of the maximum allowable temperature difference between the center and the edge of the cook machine pot;

[0036] The heating stability threshold F4 is set as the heating stability threshold F4 based on the value when the heating rate ψ of the asymmetric thermal-sensitive cell at the bottom of the cook machine reaches a stable state;

[0037] S42, based on the obtained multi-dimensional temperature control strategy threshold, classify all the local thermal anomaly correction temperature difference indexes ΔT in the correction temperature difference index set FΔT and the multi-dimensional temperature control strategy threshold, and generate a grade label L for each acquisition point i;

[0038] The grade label L is obtained through the following mechanism;

[0039]

[0040] Among them, L i represents the grade label of the i-th acquisition point, A represents the area divided into the A-level response area, B represents the area divided into the B-level response area, C represents the area divided into the C-level response area, and ψ i represents the heating rate of the i-th acquisition point.

[0041] Preferably, the S5 includes S51, S52, and S53;

[0042] S51. For the different grade labels L divided for the bottom of the cooking machine pot, allocate the basic heating power Jc corresponding to different grade labels L.

[0043] Among them, the grade label L is divided into the A-level response area, and the corresponding basic heating power Jc is 70W.

[0044] The grade label L is divided into the B-level response area, and the corresponding basic heating power Jc is 50W.

[0045] The grade label L is divided into the C-level response area, and the corresponding basic heating power Jc is 30W.

[0046] And after performing data standardization processing on the basic heating power Jc, in combination with the standard thermal perception data set, calculate and output the control power value Q of each collection point on the bottom of the cooking machine pot, and adjust the heating power of the heating module in the area of the bottom of the cooking machine pot at each collection point.

[0047] The control power value Q is calculated and output through the following algorithm formula;

[0048]

[0049] In the formula, Q i represents the control power value of the i-th collection point, and Jc Li represents the basic heating power corresponding to the grade label L of the i-th collection point, sin represents the cosine function, τ represents the current heating time, and exp represents the exponential function.

[0050] Preferably, S52. Based on the control power value Q of each collection point i combined with the standard thermal perception data set, perform summary calculation to output the comprehensive evaluation value Rk of the intelligent thermal control effect, and analyze the control quality of the temperature control and power output of the cooking machine.

[0051] The comprehensive evaluation value Rk of the intelligent thermal control effect is calculated and output through the following algorithm formula;

[0052]

[0053] In the formula, n represents the total number of collection points, and T target represents the target set temperature.

[0054] Preferably, S53. According to the mean value of the historical comprehensive evaluation value Rk of the intelligent thermal control effect during the operation of the bottom of the cooking machine pot, perform summation calculation with the standard deviation of the historical comprehensive evaluation value Rk of the intelligent thermal control effect to obtain the effect threshold Rth, and compare and evaluate the real-time obtained comprehensive evaluation value Rk of the intelligent thermal control effect with the effect threshold Rth to analyze the current temperature control strategy situation. The specific evaluation content is as follows;

[0055] When the comprehensive evaluation value Rk of the intelligent thermal control effect ≤ the effect threshold Rth, it indicates that the control effect is normal. At this time, the current power output control strategy is maintained;

[0056] When the comprehensive evaluation value Rk of the intelligent thermal control effect > the effect threshold Rth, it indicates that the current control deviates from the target and the power usage is ineffective. At this time, parameter backtracking is started, the policy level division is adjusted until the control effect is normal and the backtracking stops;

[0057] The parameter backtracking is achieved by reducing the upper temperature threshold F1 by 0.3 and increasing the medium temperature difference threshold F3 by 0.5.

[0058] An intelligent control system for a chef machine based on the Internet of Things includes a bottom pot cell division module, a cloud processing module, a local thermal anomaly analysis module, a temperature control area division module, and a temperature control backtracking module;

[0059] The bottom pot cell division module divides the bottom pot of the chef machine into asymmetric thermal-sensitive cells, sets the hottest point in the asymmetric thermal-sensitive cell as the collection point, and records the thermal perception data of each collection point based on the collection point;

[0060] The cloud processing module builds a cloud server, sets up the Internet of Things to transmit the collected thermal perception data to the cloud server, and preprocesses the thermal perception data in the cloud server to obtain a standard thermal perception data set;

[0061] The local thermal anomaly analysis module calculates and outputs the local thermal anomaly correction temperature difference index △T based on the standard thermal perception data set, traverses the local thermal anomaly correction temperature difference index △T between all pairs of collection points, and aggregates to obtain the correction temperature difference index set F△T;

[0062] The temperature control area division module divides all the local thermal anomaly correction temperature difference indexes △T in the correction temperature difference index set F△T based on the multi-dimensional temperature control strategy threshold, divides different asymmetric thermal-sensitive cells at the bottom of the chef machine into different response areas, and generates a level label L;

[0063] The temperature control backtracking module calculates and outputs the control power value Q of the heating module for each collection point respectively based on the level label L, calculates and outputs the comprehensive evaluation value Rk of the intelligent thermal control effect based on the control power value Q combined with the standard thermal perception data set, and sets the effect threshold Rth to compare and evaluate with the comprehensive evaluation value Rk of the intelligent thermal control effect to analyze the temperature control situation.

[0064] The present invention provides an intelligent control method and system for a chef machine based on the Internet of Things. It has the following beneficial effects:

[0065] (1) This method constructs a two-dimensional thermal property distribution map of the bottom of the pot based on the pot structure and manufacturing data of the cooking machine, loads a finite element simulation model of heat conduction, simulates the heat diffusion behavior of the pot body in the embedded edge computing module, and performs spatial registration and difference operation in combination with the temperature distribution map collected in real time by an infrared thermal imager, so as to realize the online self-calibration of the thermal field simulation. Subsequently, the bottom of the pot is divided into asymmetric thermosensitive cells through the region growing algorithm and the weighted Voronoi strategy, and the highest temperature point in each cell is used as the heat sensing acquisition point to extract multi-dimensional thermal sensing data including temperature values, heat capacity density, image disturbance characteristics, etc. This mechanism not only improves the modeling accuracy of the heterogeneous thermal response of the bottom of the pot, but also effectively improves the hot spot recognition and interference elimination capabilities, providing high-confidence thermal sensing basic data for subsequent refined temperature control strategies.

[0066] (2) After the thermal sensing data standardization and feature extraction are completed, the present invention constructs a standard thermal sensing data set including characteristic parameters such as heating rate and infrared image reflection disturbance factor, calculates and outputs the local thermal anomaly correction temperature difference index △T, and constructs a correction temperature difference index set F△T after traversing all pairs of thermal sensing points. Subsequently, based on the set multi-dimensional temperature control strategy threshold, each group of local thermal anomaly correction temperature difference index △T values in F△T is combined with its corresponding heating rate and divided into different response level regions, and corresponding level labels L are assigned. This division strategy pays more attention to comprehensive thermal characteristics such as heating trend, image interference and thermal inertia compared with the traditional method that only uses temperature or average difference as the judgment basis, and can effectively improve the high-temperature risk recognition accuracy and the energy efficiency control ability in the low-temperature region, so that the heating strategy is significantly improved in terms of response speed, control accuracy and energy saving effect.

[0067] (3) This method sets the basic heating power Jc according to each response level label L, constructs a non-linear heating control formula, combines parameters such as real-time heating time, heat capacity density, and image disturbance factor, calculates and outputs the control power value Q of each acquisition point, and realizes a regional differential flexible heating strategy. Subsequently, based on the temperature control results and output power of all acquisition points, an intelligent comprehensive thermal control effect evaluation value Rk is constructed and compared with the effect threshold Rth constructed based on historical statistical results to judge whether the current temperature control strategy meets the standard. If the intelligent comprehensive thermal control effect evaluation value Rk exceeds the effect threshold Rth, the parameter backtracking mechanism will be automatically executed to fine-tune the temperature upper limit threshold F1 and the medium temperature difference threshold F3, and re-divide the level labels and power output, thus constructing a thermal control closed loop with real-time self-evaluation, self-optimization and adaptive response capabilities. While ensuring the temperature stability of the bottom of the pot and the uniformity of heat energy distribution, this mechanism greatly reduces the mis-control rate and power waste, and significantly improves the cooking effect and operation energy efficiency of the whole machine. Description of the Drawings

[0068] Figure 1Schematic diagram of the steps of an intelligent control method for a chef machine based on the Internet of Things according to the present invention;

[0069] Figure 2 Schematic diagram of the process of an intelligent control system for a chef machine based on the Internet of Things according to the present invention;

[0070] Figure 3 Simulated thermal field distribution diagram according to the present invention. Detailed implementation manners

[0071] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0072] Embodiment 1

[0073] Please refer to Figure 1 and Figure 3 , the present invention provides an intelligent control method for a chef machine based on the Internet of Things. To achieve the above objectives, the present invention is realized through the following technical solutions: including the following steps:

[0074] S1. Divide the bottom of the chef machine pot into asymmetric thermosensitive cells, set the hottest point in the asymmetric thermosensitive cell as the acquisition point, and record the thermal perception data of each acquisition point based on the acquisition point;

[0075] S2. Build a cloud server, set the Internet of Things to transmit the collected thermal perception data to the cloud server, and preprocess the thermal perception data in the cloud server to obtain a standard thermal perception data set;

[0076] S3. Based on the standard thermal perception data set, calculate and output the local thermal anomaly correction temperature difference index △T, traverse the local thermal anomaly correction temperature difference index △T between all pairs of acquisition points, and then summarize to obtain the correction temperature difference index set F△T;

[0077] S4. Set the multi-dimensional temperature control strategy threshold, divide all the local thermal anomaly correction temperature difference indexes △T in the correction temperature difference index set F△T based on the multi-dimensional temperature control strategy threshold, divide different asymmetric thermosensitive cells at the bottom of the chef machine pot into different response areas, and generate a level label L;

[0078] S5. Based on the level label L, calculate and output the control power value Q of the heating module at each collection point respectively. Then, based on the control power value Q and combined with the standard thermal perception data set, calculate and output the comprehensive evaluation value Rk of the intelligent thermal control effect. Set the effect threshold Rth and compare it with the comprehensive evaluation value Rk of the intelligent thermal control effect to analyze the temperature control situation.

[0079] In this embodiment, the method loads the thermal model of the cookware structure and the real-time thermal image, online divides the asymmetric thermal-sensitive cells, extracts the local high-temperature points as the collection points, and obtains the data basis with thermal response characteristics; in step S2, the Internet of Things technology is used to transmit the real-time collected thermal perception data to the cloud server for feature extraction and standardization processing to form a standard thermal perception data set with comparability and discriminability; in S3, based on the above data set, calculate the local thermal anomaly correction temperature difference △T between the collection points, and comprehensively generate a correction index set F△T reflecting the dynamic global temperature difference; then in S4, according to the set multi-dimensional temperature control strategy threshold, complete the intelligent level division of the bottom response area of the pot to form A / B / C three-level labels; finally in S5, set the heating strategy for different level areas, calculate and output the flexibly adjusted control power Q, and at the same time construct the comprehensive evaluation value Rk of the overall intelligent thermal control effect, and compare it with the effect threshold Rth to realize the dynamic self-evaluation of performance and the closed-loop optimization of the strategy. Through the above integrated process of perception, judgment, execution and optimization, the present invention achieves multiple key technical objectives: firstly, accurately identify the real hot spots and potential thermal uneven areas at the bottom of the pot, and improve the accuracy of temperature control response; secondly, based on the heat capacity, heating rate and image disturbance factor, perform temperature difference correction calculation to improve the anti-disturbance robustness; thirdly, based on the level division mechanism, realize differential zone heating control and enhance the thermal energy utilization efficiency; fourthly, introduce the evaluation-backtracking self-adjustment mechanism to realize the online dynamic self-optimization of the temperature control strategy. In summary, the present invention not only significantly improves the temperature control accuracy and thermal control flexibility of the intelligent cooking machine, but also shows excellent comprehensive performance in aspects such as energy conservation and consumption reduction, cooking uniformity, and stability, and has broad engineering application prospects and market promotion value.

[0080] Embodiment 2

[0081] Please refer to Figure 1 and Figure 3 , specifically: S1 includes S11 and S12;

[0082] S11. According to the cookware design drawings and manufacturing process data of the cooking machine, obtain the material composition information and thickness parameters of each area at the bottom of the pot. Through the preset material database, map each area to the corresponding thermal conductivity k, specific heat capacity cp and heat capacity density value ρ, calculate the thermal diffusivity a = k / ρ·cp, and construct a thermal physical property distribution map in the two-dimensional coordinate system of the bottom of the cooking machine pot;

[0083] A two-dimensional heat conduction finite element simulation model is loaded into the embedded edge computing module inside the food processor. Based on the thermophysical property distribution map, the boundary conditions of the pot bottom and the heat source input of the heating area are set. The transient heat diffusion path of the food processor pot bottom under unit heating power is simulated to generate a simulated thermal field distribution map.

[0084] The infrared thermal imager is used to collect the temperature distribution map of the bottom of the kitchen machine pot in real time. The temperature distribution map of the bottom of the kitchen machine pot is spatially aligned and difference calculated with the simulated thermal field distribution map to obtain the thermal field difference map, and the thermal simulation model is self-calibrated online.

[0085] After the simulation's online self-calibration, the team introduced a region growing algorithm and a weighted Voronoi partitioning strategy to subdivide the high thermal gradient regions and merge and simplify the thermally stable regions. This allowed them to divide the kitchen machine base into multiple asymmetric, thermally sensitive cells. This approach ensured minimal temperature variation and high material consistency, which were the criteria for reasonable region division.

[0086] S12. Based on the asymmetric thermosensitive cells, automatically identify the highest temperature point in each asymmetric thermosensitive cell, set the highest temperature point as the collection point of the asymmetric thermosensitive cell, and then aggregate all the collection points to form a thermal sensing point set P. At the same time, each collection point records the thermal sensing data of different asymmetric thermosensitive cells in real time;

[0087] Thermal sensing data includes the temperature value T of the collection point, the heat capacity density value ρ, the local variance Varblock, the edge strength Edge and the local information entropy Entropy;

[0088] Among them, the temperature value T of the collection point is obtained in real time through infrared thermal imager;

[0089] The heat capacity density value ρ is obtained by the chef through the preset material database of the chef machine pot bottom;

[0090] The local variance Varblock is obtained by counting the variance of the pixel values in the current asymmetric thermal cell, reflecting the degree of fluctuation of the pixel grayscale value in the area. The greater the fluctuation, the more unstable the image is and the more likely it is to have steam and disturbance.

[0091] Edge intensity is obtained by using the Sobel operator to calculate the sum of gradient amplitudes, indicating whether there are abnormal edge contours in the area, such as metal reflective edges and liquid surface reflections;

[0092] Local information entropy Entropy is calculated by using the Shannon entropy formula -∑P i *log*P i Calculate and obtain, P iIt represents the i-th acquisition point where the heat-sensitive points are concentrated. "log" represents the natural logarithm function, which is used to measure the information complexity of the asymmetric thermosensitive cell. A high value indicates that the pixel distribution of the asymmetric thermosensitive cell is chaotic.

[0093] In this embodiment, in step S11 of this method, first, by analyzing the design drawings and manufacturing process data of the cook machine pot, the material properties and structural characteristics of each area of the pot bottom are extracted. Combining with the preset material database, they are mapped to the thermal conductivity k, specific heat capacity cp, and heat capacity density value ρ, and the thermal diffusivity a = k / ρ·cp is calculated to construct a high-precision thermal property distribution map in the two-dimensional coordinate system of the pot bottom. Subsequently, a two-dimensional heat conduction finite element model is loaded into the embedded edge computing module, and the boundary conditions and heat source input are set to simulate the transient heat diffusion behavior of the pot bottom under a unit heating power, and the simulated thermal field distribution map is output. Then, the simulated result is spatially registered and difference calculated with the actual temperature distribution map collected by the infrared thermal imager to form a thermal field difference map, and based on this, the thermal simulation model is corrected online to ensure that the thermal model has a high degree of fitting to the actual operating state. Finally, by introducing the region growing and weighted Voronoi partitioning strategy, the areas with sharp gradients in the thermal map are refined, and the thermally stable areas are merged and simplified, realizing the division of the asymmetric thermosensitive cells of the pot bottom. Immediately afterwards, in S12, based on the above division results, the highest temperature point in each thermosensitive cell is automatically identified and set as the acquisition point, and all acquisition points are summarized to form a heat-sensitive point set P = {P1, P2,..., Pi}. Each acquisition point records in real time multi-dimensional thermal perception data including the acquisition point temperature value T, heat capacity density value ρ, local variance Varblock, edge strength Edge, and local information entropy Entropy, etc. Among them, the temperature value is directly obtained by the infrared thermal imager, and the heat capacity density comes from the pot material library; the image perturbation-related features are extracted through image analysis algorithms to identify interference areas such as steam, reflection, and water vapor perturbation, providing an anti-interference reference basis for subsequent temperature control decisions. In summary, through the joint implementation of S11 and S12, not only is the structural subdivision of the heat response behavior of the pot bottom realized in the physical dimension, but also a multi-modal thermal perception point set is constructed in the data dimension, improving the understanding ability and response accuracy of the actual thermal field of the pot bottom.

[0094] Embodiment 3

[0095] Please refer to Figure 1 , specifically: S2 includes S21 and S22;

[0096] S21. Build a cloud server for the cook machine through the Internet of Things, and use loT Internet of Things technology to transmit the thermal perception data to the cloud server;

[0097] S22. In the cloud server, preprocess the thermal perception data to obtain a standard thermal perception data set, where the standard thermal perception data set includes the temperature value T of the collection point, the heat capacity density value ρ, the heating rate ψ, and the infrared image reflection disturbance factor φ;

[0098] The preprocessing includes data normalization processing and data feature extraction;

[0099] The data normalization processing uses Z-score normalization to perform normalization processing on the thermal perception data, eliminating the dimensionality influence between all parameters in the thermal perception data;

[0100] The data feature extraction obtains the heating rate ψ and the infrared image reflection disturbance factor φ by performing feature extraction based on the thermal perception data;

[0101] The heating rate ψ is obtained by extracting the temperature value T of the collection point and calculating the digital difference of the temperature value T of the collection point within a fixed time interval. The specific algorithm formula is: In the formula, T i (t) represents the temperature value of the i-th collection point at time t, T i (t + △t) represents the temperature value of the i-th collection point at the next (t + △t) time, △t represents the time interval, and ψ i represents the heating rate of the i-th collection point;

[0102] The infrared image reflection disturbance factor φ is obtained by performing weighted summation calculation on the local variance Varblock, edge intensity Edge, and local information entropy Entropy after the normalization processing.

[0103] In this embodiment, the method in step S2 completes the network transmission, preprocessing, and feature extraction of thermal perception data through specific operations including S21 and S22, providing standardized data support for subsequent thermal field analysis and intelligent control. In S21, first, a cloud server platform for the cook machine is constructed relying on the Internet of Things (IoT) architecture, and the multi-dimensional thermal perception data of each collection point at the bottom of the pot is stably and efficiently transmitted to the cloud server in real time through the LoT wireless communication module connected to the embedded device, realizing the linkage docking between the device side and the data side. In S22, a two-level preprocessing process is performed on the uploaded thermal perception data: one is data standardization processing, where the data at the collection points, such as temperature values and image features, are normalized through the Z-score standardization algorithm to eliminate the dimensional differences and scale inconsistencies between different parameters, thereby enhancing the convergence and comparability of the data model; the other is feature extraction operation, where two key evaluation indicators, the heating rate ψ and the infrared image reflection perturbation factor φ, are extracted respectively. Among them, the heating rate ψ is dynamically calculated in the form of numerical difference from the temperature values T of the time series collection points, reflecting the thermal change trend of each region and used to evaluate the thermal response activity degree of the region. The infrared image reflection perturbation factor φ is calculated by weighted fusion of the local variance Varblock, edge intensity Edge, and information entropy Entropy of the standardized image features to construct an index measuring the image interference intensity. The higher the value, the lower the signal-to-noise ratio of the image in that region, and it needs to be suppressed during control. Through the specific implementation of the above S2 step, the present invention not only establishes a stable cloud data perception platform but also realizes the format standardization, feature purification, and dynamic index extraction of the original multi-modal thermal data, making the thermal behaviors of each collection point comparable and providing a basis for judgment.

[0104] Embodiment 4

[0105] Please refer to Figure 1 , specifically: S3 includes S31 and S32;

[0106] S31. Based on the obtained standard thermal perception data set, calculate and output the local thermal anomaly correction temperature difference index △T between two collection points, and analyze the degree of thermal anomaly between each pair of collection points;

[0107] The local thermal anomaly correction temperature difference index △T is calculated and output through the following algorithm formula;

[0108]

[0109] In the formula, △T ij represents the local thermal anomaly correction temperature difference index between the i-th collection point and the j-th collection point, φ i represents the infrared image reflection perturbation factor of the i-th collection point, φ j represents the infrared image reflection perturbation factor of the j-th point, ρi represents the heat capacity density value of the i-th acquisition point, ρ j represents the heat capacity density value of the j-th acquisition point;

[0110] |T i -T j | represents the original temperature difference term, that is, the absolute difference in the surface temperatures of two regions, which is the most basic characterization of thermal imbalance;

[0111] represents the perturbation and heat capacity correction term. If φ i > φ j , it indicates that the acquisition point is more likely to be disturbed, such as by steam or reflection, which will weaken the temperature difference weight; ρ i + ρ j The larger it is, the stronger the thermal inertia of these two regions. Even if there is a temperature difference, it is not easy to adjust quickly, and the intervention intensity should be reduced; overall, it is a weighted correction term for perturbation response sensitivity;

[0112] The physical meaning of the formula is that the local thermal anomaly correction temperature difference index △T between the i-th acquisition point and the j-th acquisition point ij is a local thermal anomaly correction temperature difference index used to measure the effective thermal difference between two temperature acquisition points; it not only considers the original temperature difference term ∣T i -T j ∣, but also combines the heat capacity density value ρ and the infrared image reflection perturbation factor φ to perform weighted correction on the original temperature difference. If the temperature difference between two regions is large, and the image interference is small and the heat capacity is low, then the local thermal anomaly correction temperature difference index △T ij will increase, and this region will be identified as a true high-temperature difference region; if the temperature difference is large but there is image reflection or local high-heat-capacity materials, this value will be weakened to avoid misjudging as an abnormal hot spot.

[0113] S32. Traverse the standard thermal perception data set of the acquisition points in all asymmetric thermosensitive cells, calculate and output the local thermal anomaly correction temperature difference index △T of all the acquisition points in the thermal sensing point set, and generate and summarize the local thermal anomaly correction temperature difference index △T output by all the acquisition points in the thermal sensing point set to obtain the correction temperature difference index set F△T;

[0114] The expression of the correction temperature difference index set F△T is: Among them, represents a pair of points where all acquisition points i and acquisition point j belong to the thermal sensing point set P, that is, all combinations of acquisition point pairs.

[0115] In this embodiment, in step S3 of the method, through a calculation process including S31 and S32, the correction evaluation of local thermal anomaly differences and the construction of global temperature difference features are completed, providing a high-confidence criterion with anti-interference ability for subsequent response level classification. Specifically, in S31, based on the core parameters in the standard thermal perception dataset, including the temperature value T of the acquisition point, the heat capacity density value ρ, the heating rate ψ, and the infrared image reflection perturbation factor φ, etc., the local thermal anomaly correction temperature difference index ΔT between the i-th acquisition point and the j-th acquisition point is constructed ij , which is used to evaluate the effective thermal difference degree between any two acquisition points. This index not only considers the original temperature difference term but also introduces a perturbation and heat capacity correction term to eliminate the misjudgment of false hot spots caused by environmental factors such as image reflection and steam perturbation. At the same time, this correction model can also identify the regulation lag in areas with large thermal inertia, thereby reasonably reducing the intervention intensity and improving the robustness of temperature control judgment. Entering the S32 stage, further traverse all acquisition point pairs in the thermal sensing point set, and sequentially output the local thermal anomaly correction temperature difference index △T between the i-th acquisition point and the j-th acquisition point for each pair of acquisition points ij , and summarize and construct the correction temperature difference index set F△T. This set not only retains the global temperature difference topological characteristics of the bottom heat field of the pot but also forms a temperature difference structured index network for hierarchical analysis of strategies. Compared with the traditional method that relies on the average temperature difference or the maximum temperature difference for judgment, the ΔTij constructed in S3 of the present invention has higher discrimination, anti-noise ability, and physical rationality. Especially in a high-perturbation environment, it can still accurately identify the actual high thermal anomaly area, significantly reducing the misidentification rate and temperature control volatility

[0116] Embodiment 5

[0117] Please refer to Figure 1 , specifically: S4 includes S41 and S42;

[0118] S41. The multi-dimensional temperature control strategy thresholds include a temperature upper limit threshold F1, a heating rate threshold F2, a medium temperature difference threshold F3, and a heating stability threshold F4;

[0119] The temperature upper limit threshold F1 is set by the maximum allowable temperature difference between the center and the edge of the cook machine pot;

[0120] The heating rate threshold F2 is set by the upper limit value of the heating rate ψ of the cook machine in different cooking modes;

[0121] The medium temperature difference threshold F3 is set by 50% of the maximum allowable temperature difference between the center and the edge of the cook machine pot;

[0122] The heating stability threshold F4 is set as the heating stability threshold F4 based on the value of the heating rate ψ of the asymmetric thermal sensitive cells at the bottom of the cook machine when it reaches the stable state;

[0123] S42. Based on the obtained multi-dimensional temperature control strategy thresholds, classify all the local thermal anomaly correction temperature differences ΔT in the corrected temperature difference index set FΔT and the multi-dimensional temperature control strategy thresholds into levels, and generate a level label L for each acquisition point i;

[0124] The level label L is obtained through the following mechanism;

[0125]

[0126] where L i represents the level label of the i-th acquisition point, that is, L i = A means that the level label of the i-th acquisition point is divided into the A-level response area, A means divided into the A-level response area, B means divided into the B-level response area, C means divided into the C-level response area, and ψ i represents the heating rate of the i-th acquisition point;

[0127] △T ij > F1 ∧ ψ i > F2 means that when the local temperature of the current acquisition point area is significantly higher than the surrounding area and the heating rate is fast, it is divided into the A-level response area;

[0128] F3 < △T ij ≤ F1 means that there is a temperature difference at the current acquisition point i but it is not severe, the temperature change is stable or gently rising, and it is divided into the B-level response area;

[0129] △T ij ≤ F3 ∧ ψ i < F4 means that the temperature difference is small, the heating rate is slow or almost stationary, and it is divided into the C-level response area.

[0130] In this embodiment, step S4 of this method completes the hierarchical mapping from the thermal difference characteristics to the response strategy through two sub-processes, S41 and S42, realizing the structuring and intelligentization of the bottom-of-pot thermal control decision-making. In S41, first, multi-dimensional temperature control strategy thresholds are set, specifically including the temperature upper limit threshold F1, the heating rate threshold F2, the medium temperature difference threshold F3, and the heating stability threshold F4. This setting mechanism ensures that the granularity and determination logic of the temperature control strategy have physical significance and engineering adaptability, providing a multi-dimensional threshold basis for matching different thermal control strategies for different regions. In S42, the corrected temperature difference index set FΔT constructed in the previous stage is compared and analyzed item by item with the above multi-dimensional thresholds, and combined with the heating rate ψ of each acquisition point i, the response level division strategy is executed to generate a level label set L = {Li}. This mechanism implements the response level judgment logic of the "temperature difference range + dynamic thermal response" dual factor, enabling the temperature control strategy to not only focus on the temperature itself, but also on its change trend and regional stability, avoiding misjudgment responses to temporary or disturbing hot spots. Through the implementation of S4, the present invention effectively establishes a bridge between thermal sensing data and temperature control strategies, achieving precise, dynamic, and differentiated control of the bottom heating behavior of the pot. Compared with the traditional method based on fixed temperature difference or single threshold control, this method can adopt customized responses for different regions of the bottom of the pot according to the thermal evolution behavior, improving the flexibility, robustness, and energy efficiency utilization rate of temperature control. It is particularly suitable for cooking scenarios with multi-material structure cookware and complex working conditions, significantly enhancing the intelligent level of the cooking machine and the user cooking experience.

[0131] Example 6

[0132] Please refer to Figure 1 , specifically: S5 includes S51, S52, and S53;

[0133] S51. For the different level labels L divided for the bottom of the cooking machine pot, different basic heating powers Jc corresponding to the level labels L are allocated.

[0134] Among them, the level label L is divided into the A-level response area, and the corresponding basic heating power Jc is 70W;

[0135] The level label L is divided into the B-level response area, and the corresponding basic heating power Jc is 50W;

[0136] The level label L is divided into the C-level response area, and the corresponding basic heating power Jc is 30W;

[0137] After data standardization processing of the basic heating power Jc and in combination with the standard thermal perception data set, the control power value Q of each collection point on the bottom of the cooking machine pot is calculated and output, and the heating power of the heating module in the bottom area of the cooking machine pot at each collection point is adjusted;

[0138] The control power value Q is calculated and output through the following algorithm formula;

[0139]

[0140] In the formula, Q i represents the control power value of the i-th collection point, and Jc Li represents the basic heating power corresponding to the level label L of the i-th collection point, sin represents the cosine function, τ represents the current heating time, and exp represents the exponential function;

[0141] sin(ψ i ·ρ i·τ) represents the thermal response fluctuation term. This sine function simulates the thermal inertia hysteresis behavior of the bottom material of the pot. The sine wave form can bring perturbation fluctuations to prevent sudden changes in control rigidity. The larger the heat capacity, the faster the temperature rise, that is, the higher the adjustment frequency and the more significant the change;

[0142] represents the image perturbation suppression factor, analyzing whether the infrared image in this area is disturbed by steam, water droplets, reflection, etc. The heating rate ψ of the i-th acquisition point in the denominator i , if the temperature rises quickly, it means that the perturbation can still be tolerated moderately to prevent overcorrection; the exponential form can quickly suppress the power in the high-perturbation area to be close to 0;

[0143] Specific example, in area A, the temperature rises violently but the image is stable;

[0144] ψ i = 2.0, ρ i = 0.8, φ i = 0.1, τ = 10;

[0145] Q i = 0.7·(1 + sin(2·0.8·10))·exp(-0.01 / 4) = 0.4972.

[0146] S52. Based on the control power value Q of each acquisition point i and combined with the standard thermal perception data set, perform summary calculations to output the comprehensive evaluation value Rk of the intelligent thermal control effect, and analyze the control quality of the temperature control and power output of the cooking machine;

[0147] The comprehensive evaluation value Rk of the intelligent thermal control effect is calculated and output through the following algorithm formula;

[0148]

[0149] In the formula, n represents the total number of acquisition points, T target represents the target set temperature, with a dimensionless value, input by the user's control target,

[0150] represents the temperature control deviation degree. The numerator represents the square of the deviation between the actual temperature and the target temperature, used to measure the temperature control accuracy. The denominator represents the heating rate multiplied by the heat capacity, used to represent the sensitivity of this area to temperature control intervention;

[0151] represents the energy control efficiency, that is, the relative energy consumption. The numerator represents the square of the current output power. The higher the energy consumption, the greater the penalty. The denominator represents the reference power that the grade label L should have.

[0152] S53. According to the mean value of the comprehensive evaluation value Rk of the historical intelligent thermal control effect during the operation of the bottom of the cooking machine, calculate the sum with the standard deviation of the comprehensive evaluation value Rk of the historical intelligent thermal control effect to obtain the effect threshold Rth. Then, compare and evaluate the comprehensive evaluation value Rk of the real-time obtained intelligent thermal control effect with the effect threshold Rth to analyze the current temperature control strategy. The specific evaluation content is as follows;

[0153] When the comprehensive evaluation value Rk of the intelligent thermal control effect ≤ the effect threshold Rth, it indicates that the control effect is normal. At this time, maintain the current power output control strategy;

[0154] When the comprehensive evaluation value Rk of the intelligent thermal control effect > the effect threshold Rth, it indicates that the current control deviates from the target and the power usage is ineffective. At this time, start parameter backtracking, adjust the strategy level division until the control effect is normal and stop backtracking;

[0155] The parameter backtracking is achieved by reducing the upper temperature threshold F1 by 0.3 and increasing the medium temperature difference threshold F3 by 0.5;

[0156] Specific example: Assume that the current comprehensive evaluation value Rk of the intelligent thermal control effect = 105 > the effect threshold Rth = 90, and the original parameters are F1 = 8.0 and F3 = 4.0. After backtracking, F1 = 7.7, F2 = 4.5, the new parameters take effect immediately, re-classify, and then re-output the basic heating power Jc corresponding to the level label L of the i-th collection point Li and the comprehensive evaluation value Rk of the intelligent thermal control effect.

[0157] In this embodiment, step S5 of this method includes three sub-processes S51, S52, and S53, constructing a complete closed-loop control mechanism from the response level label to power distribution, then to comprehensive evaluation and feedback optimization. In S51, according to the response level label L output in the previous stage S4 i , divide each thermosensitive cell at the bottom of the pot into three grade response areas A, B, and C, and assign corresponding basic heating powers Jc respectively, specifically: grade A is 70W, grade B is 50W, and grade C is 30W. On this basis, by introducing a standard thermal perception data set, use a non-linear control formula including a sine wave simulation term and an exponential suppression term to output the control power value Q of the collection point i , realizing flexible dynamic regulation of the heating modules at each collection point at the bottom of the pot. This model takes into account the thermal inertia of the pot body, the temperature rise delay fluctuation, and the image interference suppression, significantly improving the flexibility and robustness of the temperature control behavior. Entering the S52 stage, the control power values Q of all collection points iBased on temperature data, a comprehensive intelligent thermal control effectiveness evaluation function Rk is constructed. Its evaluation logic includes two core dimensions: temperature control deviation (the ratio of the deviation between the actual and target temperatures to the thermal response sensitivity); and energy control efficiency (the ratio of power usage intensity to the graded benchmark power). These two dimensions are weighted and synthesized to form a comprehensive numerical criterion for the quality of the current temperature control behavior. This metric not only reflects energy consumption but also measures whether the temperature control strategy aligns with the actual thermal response. In S53, the evaluation threshold Rth is calculated based on the historical statistics of the intelligent thermal control effectiveness comprehensive evaluation values Rk during operation, and the current evaluation value is compared with it. The grade labeling and power control processes are then re-executed to ensure that the temperature control deviation is corrected and convergence is achieved, ultimately forming a self-feedback, self-calibration, and adaptive closed-loop temperature control process. Through the integrated implementation of S5, the present invention achieves a full-cycle intelligent thermal control closed-loop, from judgment to execution, from execution to evaluation, and from evaluation to strategy self-optimization, significantly improving the temperature control stability and energy efficiency of the chef machine in complex cooking environments. This mechanism can effectively reduce temperature control deviation problems caused by strategy rigidity, uncontrolled disturbances or misjudgment of thermal inertia, ensuring a more uniform heat distribution in the bottom area of the pot and a more stable cooking effect. It has extremely high practical value and promotion prospects.

[0158] Example 7

[0159] See also Figure 2 , an intelligent control system for a chef machine based on the Internet of Things, including a pot bottom cell division module, a cloud processing module, a local thermal anomaly analysis module, a temperature control area division module and a temperature control backtracking module;

[0160] The pot bottom cell division module divides the chef machine pot bottom into asymmetric heat-sensitive cells, sets the hottest point in the asymmetric heat-sensitive cells as a collection point, and records the thermal sensing data of each collection point based on the collection point;

[0161] The cloud processing module constructs a cloud server and sets up the Internet of Things to transmit the collected thermal sensing data to the cloud server, and pre-processes the thermal sensing data in the cloud server to obtain a standard thermal sensing data set;

[0162] The local thermal anomaly analysis module calculates and outputs the local thermal anomaly corrected temperature difference index △T based on the standard thermal perception data set, and traverses the local thermal anomaly corrected temperature difference index △T between all two collection points, and then summarizes them to obtain the corrected temperature difference index set F△T;

[0163] The temperature control area division module divides all local thermal anomaly corrected temperature difference indicators ΔT in the corrected temperature difference index set FΔT into different response areas by setting multi-dimensional temperature control strategy thresholds, divides different asymmetric thermal-sensitive cells at the bottom of the cooking machine pot into different response areas, and generates a level label L;

[0164] The temperature control backtracking module calculates and outputs the control power value Q of the heating module at each acquisition point based on the level label L, calculates and outputs the comprehensive evaluation value Rk of the intelligent thermal control effect based on the control power value Q combined with the standard thermal perception data set, and sets the effect threshold Rth to compare and evaluate with the comprehensive evaluation value Rk of the intelligent thermal control effect to analyze the temperature control situation.

[0165] Although the embodiments of the present invention have been shown and described, those of ordinary skill in the art can understand that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the present invention.

Claims

1. An intelligent control method for a chef machine based on the Internet of Things, characterized in that: Including the following steps: S1. Divide the bottom of the cooking machine pot into asymmetric thermal-sensitive cells, set the hottest point in the asymmetric thermal-sensitive cell as the acquisition point, and record the thermal perception data of each acquisition point based on the acquisition point; S2. Build a cloud server, set up the Internet of Things to transmit the collected thermal perception data to the cloud server, and preprocess the thermal perception data in the cloud server to obtain a standard thermal perception data set; S3. Based on the standard thermal perception data set, calculate and output the local thermal anomaly correction temperature difference index △T, traverse the local thermal anomaly correction temperature difference index △T between all pairs of acquisition points, and then summarize to obtain the correction temperature difference index set F△T; S4. Set the multi-dimensional temperature control strategy threshold, divide all the local thermal anomaly correction temperature difference indexes △T in the correction temperature difference index set F△T based on the multi-dimensional temperature control strategy threshold, divide different asymmetric thermal-sensitive cells at the bottom of the cooking machine pot into different response areas, and generate a level label L; S5. Based on the level label L, calculate and output the control power value Q of the heating module for each acquisition point respectively, calculate and output the comprehensive evaluation value Rk of the intelligent thermal control effect based on the control power value Q combined with the standard thermal perception data set, set the effect threshold Rth and compare and evaluate it with the comprehensive evaluation value Rk of the intelligent thermal control effect to analyze the temperature control situation.

2. The intelligent control method of a chef machine based on the Internet of Things according to claim 1, wherein: The S1 includes S11 and S12; S11. According to the design drawings and manufacturing process data of the cooking machine pot, obtain the material composition information and thickness parameters of each area at the bottom of the pot, map each area to the corresponding thermal conductivity k, specific heat capacity cp and heat capacity density value ρ through a preset material database, calculate and obtain the thermal diffusivity a = k / ρ·cp, and build a thermal physical property distribution map in the coordinate system of the two-dimensional bottom of the cooking machine pot; In the embedded edge computing module inside the cooking machine, load a two-dimensional heat conduction finite element simulation model, set the bottom boundary conditions and heat source input in the heating area based on the thermal physical property distribution map, simulate the transient heat diffusion path of the bottom of the cooking machine pot under the unit heating power, and generate a simulated thermal field distribution map; Real-time collect the temperature distribution map of the bottom of the cooking machine pot through an infrared thermal imager, perform spatial registration and difference operation on the temperature distribution map of the bottom of the cooking machine pot and the simulated thermal field distribution map to obtain a thermal field difference map, and perform online self-calibration on the thermal simulation model; After the online self-calibration of the simulation, introduce a region growing algorithm and a weighted Voronoi partitioning strategy to subdivide the grid in the high thermal gradient area, merge and simplify the area with stable residual heat, and divide the bottom of the cooking machine pot into multiple asymmetric thermal-sensitive cells; S12. Based on the asymmetric thermal-sensitive cells, automatically identify the highest temperature point in each asymmetric thermal-sensitive cell, set the highest temperature point as the acquisition point of the asymmetric thermal-sensitive cell, then summarize all the acquisition points to obtain a thermal perception point set P, and at the same time, each acquisition point records the thermal perception data in different asymmetric thermal-sensitive cells in real time; The thermal perception data includes the acquisition point temperature value T, heat capacity density value ρ, local variance Varblock, edge strength Edge and local information entropy Entropy.

3. The intelligent control method of a planetary mixer based on the Internet of Things according to claim 2, characterized in that: The S2 includes S21 and S22; S21. Build a cloud server for the cooking machine through the Internet of Things, and use IoT technology to transmit thermal perception data to the cloud server; S22. In the cloud server, preprocess the thermal perception data to obtain a standard thermal perception data set, where the standard thermal perception data set includes the temperature value T at the collection point, the heat capacity density value ρ, the heating rate ψ, and the infrared image reflection disturbance factor φ; The preprocessing includes data standardization processing and data feature extraction; The data standardization processing performs normalization standardization processing on the thermal perception data by using Z-score standardization to eliminate the dimensional influence between all parameters in the thermal perception data; The data feature extraction obtains the heating rate ψ and the infrared image reflection disturbance factor φ by performing feature extraction based on the thermal perception data.

4. The intelligent control method of a chef machine based on the Internet of Things according to claim 3, characterized in that: The S3 includes S31 and S32; S31. Based on the obtained standard thermal perception data set, calculate and output the local thermal anomaly correction temperature difference index △T between two collection points, and analyze the degree of thermal anomaly between each pair of collection points; The local thermal anomaly correction temperature difference index △T is calculated and output through the following algorithm formula; Where, △T ij represents the local thermal anomaly correction temperature difference index between the i-th acquisition point and the j-th acquisition point, φ i represents the infrared image reflection disturbance factor of the i-th acquisition point, φ j represents the infrared image reflection disturbance factor of the j-th point, ρ i represents the heat capacity density value of the i-th acquisition point, ρ j represents the heat capacity density value of the j-th acquisition point.

5. The intelligent control method of a chef machine based on the Internet of Things according to claim 4, characterized in that: S32. Traverse the standard thermal perception data set of the collection points in all asymmetric thermal-sensitive cells, calculate and output the local thermal anomaly correction temperature difference index △T of the collection points in all thermal sensing point sets, and generate and summarize the local thermal anomaly correction temperature difference index △T output by the collection points in all thermal sensing point sets to obtain the correction temperature difference index set F△T; The expression of the corrected temperature difference index set F△T is as follows: where indicates that all the collection points i and the collection point j are a pair of points belonging to the thermal sensation point set P.

6. The intelligent control method of a chef machine based on the Internet of Things according to claim 1, wherein: The S4 includes S41 and S42; S41. The multi-dimensional temperature control strategy thresholds include a temperature upper limit threshold F1, a heating rate threshold F2, a medium temperature difference threshold F3, and a heating stability threshold F4; The temperature upper limit threshold F1 is set by the maximum allowable temperature difference between the center and the edge of the cooking machine pot; The heating rate threshold F2 is set by the upper limit value of the heating rate ψ of the cooking machine at low temperature under different cooking modes; The medium temperature difference threshold F3 is set by 50% of the maximum allowable temperature difference between the center and the edge of the cooking machine pot; The heating stability threshold F4 is set as the heating stability threshold F4 based on the value when the heating rate ψ of the asymmetric thermal-sensitive cells at the bottom of the cooking machine reaches a stable state; S42. Based on the obtained multi-dimensional temperature control strategy thresholds, classify all the local thermal anomaly correction temperature difference indexes △T in the correction temperature difference index set F△T and the multi-dimensional temperature control strategy thresholds, and generate a grade label L for each collection point i; The grade label L is obtained through the following mechanism; Among them, L i represents the level label of the i-th collection point, A represents being divided into a Class A response area, B represents being divided into a Class B response area, C represents being divided into a Class C response area, ψ i represents the heating rate of the i-th collection point.

7. The intelligent control method of a chef machine based on the Internet of Things according to claim 6, wherein: The S5 includes S51, S52, and S53; S51. For the different grade labels L divided for the bottom of the cooking machine, allocate the basic heating power Jc corresponding to the different grade labels L, where the grade label L is divided into the A-level response area, and the corresponding basic heating power Jc is 70W; The grade label L is divided into the B-level response area, and the corresponding basic heating power Jc is 50W; The grade label L is divided into the C-level response area, and the corresponding basic heating power Jc is 30W; After normalizing the data of the basic heating power Jc and combining it with the standard thermal perception dataset, the control power value Q of each collection point on the bottom of the cooking machine pot is calculated and output, and the heating power of the heating module in the bottom area of the cooking machine pot at each collection point is adjusted; The control power value Q is calculated and output through the following algorithm formula; Where, Q i represents the control power value of the i-th acquisition point, and Jc Li represents the basic heating power corresponding to the grade label L of the i-th acquisition point. sin represents the cosine function, τ represents the current heating time, and exp represents the exponential function.

8. The intelligent control method of a chef machine based on the Internet of Things according to claim 6, characterized in that: S52. Based on the control power value Q of each collection point i and combining the standard thermal perception dataset, the comprehensive evaluation value Rk of the intelligent thermal control effect is calculated and output through summary calculation, and the control quality of the temperature control and power output of the cooking machine is analyzed; The comprehensive evaluation value Rk of the intelligent thermal control effect is calculated and output through the following algorithm formula; Where n represents the total number of acquisition points, and T target represents the target set temperature.

9. The intelligent control method of a chef machine based on the Internet of Things according to claim 8, wherein: S53. According to the mean value of the historical comprehensive evaluation value Rk of the intelligent thermal control effect during the operation of the bottom of the cooking machine pot, the sum is calculated with the standard deviation of the historical comprehensive evaluation value Rk of the intelligent thermal control effect to obtain the effect threshold Rth. Then, the real-time obtained comprehensive evaluation value Rk of the intelligent thermal control effect is compared with the effect threshold Rth to analyze the current temperature control strategy. The specific evaluation content is as follows; When the comprehensive evaluation value Rk of the intelligent thermal control effect ≤ the effect threshold Rth, it means that the control effect is normal, and the current power output control strategy is maintained at this time; When the comprehensive evaluation value Rk of the intelligent thermal control effect > the effect threshold Rth, it means that the current control deviates from the target and the power usage is invalid. At this time, parameter backtracking is started, the strategy level division is adjusted until the control effect is normal and the backtracking stops; The parameter backtracking is to lower the temperature upper limit threshold F1 by 0.3 and increase the medium temperature difference threshold F3 by 0.

5.

10. An intelligent control system for a chef machine based on the Internet of Things, which is applied to an intelligent control method for a chef machine based on the Internet of Things according to any one of claims 1-9, and is characterized in that: It includes a bottom pot cell division module, a cloud processing module, a local thermal anomaly analysis module, a temperature control area division module, and a temperature control backtracking module; The bottom pot cell division module divides the bottom of the cooking machine pot into asymmetric thermal-sensitive cells, sets the hottest point in the asymmetric thermal-sensitive cell as the collection point, and records the thermal perception data of each collection point based on the collection point; The cloud processing module builds a cloud server, sets up the Internet of Things to transmit the collected thermal perception data to the cloud server, and preprocesses the thermal perception data in the cloud server to obtain the standard thermal perception dataset; The local thermal anomaly analysis module calculates and outputs the local thermal anomaly correction temperature difference index △T based on the standard thermal perception dataset, traverses the local thermal anomaly correction temperature difference index △T between all pairs of collection points, and summarizes to obtain the correction temperature difference index set F△T; The temperature control area division module divides all the local thermal anomaly correction temperature difference indexes △T in the correction temperature difference index set F△T based on the multi-dimensional temperature control strategy threshold, divides different asymmetric thermal-sensitive cells at the bottom of the cooking machine pot into different response areas, and generates a level label L; The temperature control backtracking module calculates and outputs the control power value Q of the heating module at each acquisition point based on the level label L. Then, based on the control power value Q and combined with the standard thermal perception data set, it calculates and outputs the comprehensive evaluation value Rk of the intelligent thermal control effect. And it sets the effect threshold Rth to compare and evaluate with the comprehensive evaluation value Rk of the intelligent thermal control effect to analyze the temperature control situation.

Images