SiC substrate-based intelligent temperature control heating platform and chip processing method

Abstract

The application relates to the technical field of chip processing, and discloses an intelligent temperature control heating platform based on a SiC substrate and a chip processing method.The method comprises the following steps: modeling the temperature distribution characteristics of the SiC substrate to obtain a three-dimensional temperature field distribution model; performing temperature parameter analysis to obtain a SiC substrate surface temperature distribution parameter set; performing thermal zone analysis to obtain a heating unit contribution degree coefficient matrix and a target thermal zone distribution map; calculating a power distribution value to obtain a heating power setting matrix; performing four-stage hierarchical control on the heating power setting matrix to obtain a smooth transition control curve; and implementing a progressive temperature control strategy to obtain global temperature control parameters meeting the chip processing precision requirements of the SiC substrate.The application realizes smooth transition from room temperature to process temperature, does not need additional preheating equipment, provides a stable processing temperature environment for the SiC substrate, reduces energy consumption, and realizes energy efficient utilization in the chip processing process of the SiC substrate.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of chip processing, and in particular to an intelligent temperature control heating platform based on a SiC substrate and a chip processing method. BACKGROUND

[0002] In the process of SiC substrate chip processing, temperature control faces unique challenges. Due to the high thermal conductivity and heat capacity characteristics of SiC material, traditional silicon-based chip processing temperature control methods are difficult to adapt to the temperature characteristics of SiC substrate, resulting in problems such as uneven thermal field, excessive temperature gradient, and serious temperature fluctuation. These problems directly affect the quality of chip processing, reduce product yield, and increase manufacturing cost.

[0003] The widely used temperature control system is mainly based on single-point or multi-point PID control. These methods have obvious limitations in dealing with the complex thermal field distribution of SiC substrate. There is usually a significant temperature gradient on the surface of SiC substrate and different processing areas. The traditional control method is difficult to accurately adjust the heating power distribution, resulting in the difficulty of eliminating the influence of uneven heat on the chip surface and edge. At the same time, temperature fluctuations are easily introduced into the control system, which further affects the processing quality. In addition, the initial heating stage in the process of SiC substrate chip processing usually requires the use of additional preheating equipment, which increases energy consumption and equipment cost. SUMMARY

[0004] The present application provides an intelligent temperature control heating platform based on a SiC substrate and a chip processing method. The present application realizes smooth transition from room temperature to process temperature without the need for additional preheating equipment, provides a stable processing temperature environment for SiC substrate, reduces energy consumption, and realizes energy-efficient utilization of SiC substrate chip processing.

[0005] In the first aspect, the present application provides an intelligent temperature control heating chip processing method based on a SiC substrate, which comprises:

[0006] Modeling the temperature distribution characteristics of the SiC substrate to obtain a three-dimensional temperature field distribution model;

[0007] Performing temperature parameter analysis based on the three-dimensional temperature field distribution model to obtain a set of SiC substrate surface temperature distribution parameters;

[0008] Performing thermal zone analysis on the set of SiC substrate surface temperature distribution parameters to obtain a heating unit contribution degree coefficient matrix and a target thermal zone distribution map;

[0009] Calculating power distribution values based on the heating unit contribution degree coefficient matrix and the target thermal zone distribution map to obtain a heating power setting matrix;

[0010] Perform four-stage hierarchical control on the heating power setting matrix to obtain a smooth transition control curve;

[0011] Implement a progressive temperature control strategy according to the smooth transition control curve to obtain global temperature control parameters that meet the SiC substrate chip processing precision requirements.

[0012] In a second aspect, the present application provides a SiC substrate-based intelligent temperature control heating chip processing platform, which comprises:

[0013] A modeling module for modeling SiC substrate temperature distribution characteristics to obtain a three-dimensional temperature field distribution model;

[0014] A temperature parameter analysis module for performing temperature parameter analysis based on the three-dimensional temperature field distribution model to obtain a set of SiC substrate surface temperature distribution parameters;

[0015] A hot zone analysis module for performing hot zone analysis on the set of SiC substrate surface temperature distribution parameters to obtain a heating unit contribution degree coefficient matrix and a target hot zone distribution map;

[0016] A calculation module for calculating power distribution values based on the heating unit contribution degree coefficient matrix and the target hot zone distribution map to obtain a heating power setting matrix;

[0017] A hierarchical control module for performing four-stage hierarchical control on the heating power setting matrix to obtain a smooth transition control curve;

[0018] A temperature control module for implementing a progressive temperature control strategy according to the smooth transition control curve to obtain global temperature control parameters that meet the SiC substrate chip processing precision requirements.

[0019] In the technical scheme provided by the application, by establishing a high-precision three-dimensional temperature field distribution model, real-time temperature prediction is realized, and reliable data support is provided for accurate temperature control in SiC substrate chip processing. Based on a specific temperature parameter analysis process, target parameters such as hot spot position coordinates, average temperature value and maximum temperature gradient can be accurately obtained, and the uncertainty thereof can be quantified, laying a foundation for thermal zone analysis. By numbering and positioning the heating units and applying differentiated weight coefficients, the target heating unit that has a decisive influence on temperature uniformity is accurately identified, forming a heating unit contribution coefficient matrix and a target thermal zone distribution map. Based on this, probability-based power distribution is realized, and the target heating unit is preferentially allocated adjustment resources, while eliminating mutual interference between adjacent heating units to prevent power mutation. A four-stage hierarchical control method is adopted to realize smooth transition from room temperature to process temperature, without the need for additional preheating equipment, to provide a stable processing temperature environment for the SiC substrate. The progressive temperature control strategy divides the SiC substrate surface into three control levels, dynamically allocates control resources using a space-time evolution control model, combines adaptive parameter optimization and a multi-level safety guarantee system, and ensures high precision and high stability of temperature control. In addition, by accurately controlling the temperature distribution of the SiC substrate surface, defects caused by uneven and fluctuating temperature are significantly reduced, improving the yield and consistency of SiC chip manufacturing, while the optimized power distribution strategy also reduces energy consumption, realizing energy-efficient utilization of the SiC substrate chip processing process. BRIEF DESCRIPTION OF DRAWINGS

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

[0021] Figure 1 The flowchart of the SiC substrate-based intelligent temperature control heating chip processing method provided by the embodiments of the present application is shown.

[0022] Figure 2 The structural schematic block diagram of the SiC substrate-based intelligent temperature control heating chip processing platform provided by the embodiments of the present application is shown. DETAILED DESCRIPTION

[0023] The technical solutions in the embodiments of the present application will be described in detail below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are some embodiments of the present application, not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0024] The flowcharts shown in the drawings are merely illustrative and do not necessarily include all contents and operations / steps, nor are they necessarily executed in the order described. For example, some operations / steps can be further decomposed, combined or partially merged, so the actual execution order can be changed based on actual conditions.

[0025] It should also be understood that the terms used in this specification of the present application are only for the purpose of describing specific embodiments and are not intended to limit the present application. As used in this specification and the appended claims, the singular forms "a," "an" and "the" are intended to include plural forms unless the context clearly indicates otherwise.

[0026] It should be further understood that the term "and / or" used in this 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.

[0027] 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.

[0028] Please refer to Figure 1 , Figure 1 The flowchart of the chip processing method based on SiC substrate intelligent temperature control heating provided by the embodiments of the present application is shown in Figure 1 The chip processing method based on SiC substrate intelligent temperature control heating provided by the embodiments of the present application includes steps S100 to S600.

[0029] Step S100, model the temperature distribution characteristics of the SiC substrate to obtain a three-dimensional temperature field distribution model;

[0030] It can be understood that the execution subject of the present application can be a chip processing platform based on SiC substrate intelligent temperature control heating, and can also be a terminal or a server, and the specific implementation is not limited here. The embodiments of the present application take the server as the execution subject for example.

[0031] Specifically, the key thermophysical parameters of SiC substrate material under different temperature conditions, such as thermal conductivity, specific heat capacity and thermal diffusivity, are collected, and through a large number of experiments, the thermophysical property data of SiC substrate material under different thicknesses (0.5mm-2.0mm) and different sizes (4 inches-12 inches) are obtained, which form a thermophysical property matrix. Based on the thermophysical property matrix, the thermophysical parameter set of SiC substrate under different temperatures is generated through interpolation method or data fitting technology, reflecting the change of thermophysical properties of SiC substrate caused by temperature change in the processing process. The thermophysical parameter set and the geometric structure data of SiC substrate are input into the finite element analysis system to construct an initial three-dimensional temperature field model. The initial model includes the thickness, size and material parameters of SiC substrate, and also contains the position, shape and thermal contact impedance of the heater and other information, which fully simulates the temperature distribution of SiC substrate under different process environments. Since the temperature field of SiC substrate is affected by multiple factors such as edge heat dissipation, environmental temperature and heat source power distribution, boundary conditions are applied to the initial three-dimensional model after the initial model is constructed, including considering the heat dissipation effect of the edge area, environmental temperature change and thermal contact impedance between the heating unit and SiC substrate, to obtain the temperature field response equation. The temperature field response equation is a partial differential equation set based on heat conduction theory, which describes the temperature change state of SiC substrate at different positions and different times. Based on the temperature field response equation, real-time temperature prediction is realized, and by taking the heating power distribution matrix and the current processing conditions as input, the temperature distribution map and the temperature gradient distribution map of SiC substrate surface are obtained by solving the equation, reflecting the temperature change law of SiC substrate surface and different positions inside. Since there is a certain deviation between the theoretical prediction of the finite element model and the actual processing process, the system is calibrated by comparing the temperature distribution map and the temperature gradient distribution map with the measured temperature data. By setting thermocouples or infrared thermometers and other temperature measuring devices, the measured temperature values of SiC substrate at different positions are obtained, and these data are compared with the calculation results of the finite element model for analysis, the correction coefficient is calculated, and the model parameters are adjusted through iteration to optimize the prediction accuracy of the model, and the three-dimensional temperature field distribution model is obtained.

[0032] Step S200, performing temperature parameter analysis based on the three-dimensional temperature field distribution model to obtain a temperature distribution parameter set of the SiC substrate surface;

[0033] Specifically, the heating power distribution matrix of the SiC substrate and the environmental parameters are input into a three-dimensional temperature field distribution model, and the original temperature distribution data of the SiC substrate at different positions are obtained through finite element solution or thermal response simulation. These data contain temperature values at different time and space points, forming a temperature field information matrix. The original temperature distribution data is subjected to feature extraction and noise reduction processing. Feature extraction uses principal component analysis or wavelet transform method to map high-dimensional temperature data to low-dimensional feature space to retain key temperature distribution characteristics. At the same time, noise reduction processing eliminates noise signals through median filtering or Kalman filtering algorithm to obtain a smoother and more accurate temperature distribution feature matrix. Then, based on the temperature distribution feature matrix, the set of coordinates of hot spots on the surface of the SiC substrate is calculated. Hot spot recognition uses a temperature gradient maximum positioning algorithm to mark the area in the temperature distribution feature matrix that exceeds the set threshold as a hot spot area, and extract its coordinate position to generate the set of coordinates of surface hot spots. Temperature gradient analysis is performed on the set of coordinates of surface hot spots. The temperature change rate of the hot spot area is quantified by calculating the size and direction of the temperature gradient of the hot spot adjacent area, and the surface temperature change rate data of the SiC substrate is obtained, reflecting the change trend of temperature at different positions over time. The surface temperature change rate data is compared and analyzed with the process temperature threshold, by comparing the actual temperature change rate with the allowed temperature change rate, the deviation between the current temperature distribution of the SiC substrate and the process requirement is evaluated, and the temperature deviation evaluation result is obtained. According to the temperature deviation evaluation result, the SiC substrate surface temperature distribution parameter set containing hot spot coordinates, average temperature value, maximum temperature gradient and uncertainty quantization value is generated. The hot spot coordinates reflect the position of the highest temperature area, while the average temperature value quantifies the thermal uniformity of the SiC substrate as a whole, the maximum temperature gradient represents the temperature change amplitude at different positions on the surface of the SiC substrate, and the uncertainty quantization value provides the uncertainty range of temperature parameter prediction through heteroscedasticity estimation method.

[0034] Step S300, performing thermal zone analysis on the SiC substrate surface temperature distribution parameter set to obtain a heating unit contribution coefficient matrix and a target thermal zone distribution map;

[0035] Specifically, all heating units in the SiC substrate heating system are numbered and positioned to form a heating unit index matrix. The matrix corresponds the spatial position of the heating unit to the number, and forms a two-dimensional matrix structure according to the arrangement of the heating unit in the x direction and the y direction. The temperature uniformity index is calculated based on the hot spot coordinates and the average temperature value in the SiC substrate surface temperature distribution parameter set. The temperature uniformity index is used to quantify the distribution of the SiC substrate surface temperature. The closer the value is to the ideal state, the more uniform the temperature distribution is. When the temperature uniformity index is low, there is a large temperature gradient or hot spot area on the SiC substrate surface. These hot spot areas need to be focused on and power adjusted in the subsequent temperature control process to maintain temperature uniformity. The disturbance range of each heating unit is determined according to the maximum temperature gradient and the uncertainty quantification value in the SiC substrate surface temperature distribution parameter set. By slightly disturbing the power of the heating unit, the temperature change trend of the SiC substrate surface under different power levels is observed, and the control ability of each heating unit on the temperature uniformity and hot spot area is evaluated combined with these data. Through multiple disturbances and result analysis, a data set containing the influence range of each heating unit on the temperature distribution is formed. The hot spot coordinates in the SiC substrate surface temperature distribution parameter set are spatially mapped to the heating unit index matrix to obtain the target heating unit set. By calculating the distance relationship between the hot spot area and the heating unit, it is determined which heating units have a direct impact on the hot spot area, and these heating units are classified into the target heating unit set. According to the temperature uniformity index and the disturbance range of each heating unit, different weight factors are set for the target heating unit set and the heating units in the non-target area to obtain the heating unit contribution coefficient matrix. For the target heating unit set, a higher weight factor is given because of its greater impact on the hot spot area, while for the heating units in the non-target area, a lower weight factor is given because of their smaller impact on the temperature distribution. According to the heating unit contribution coefficient matrix, the heat zone contribution heat map is analyzed, and the hot spot coordinates in the SiC substrate surface temperature distribution parameter set are marked in the heat map to generate a target heat zone distribution map, which shows the influence degree of different heating units on the SiC substrate surface temperature distribution and highlights the position of the hot spot area.

[0036] Step S400, calculate the power allocation value based on the heating unit contribution coefficient matrix and the target heat zone distribution map to obtain a heating power setting matrix;

[0037] Specifically, the target value of temperature control is determined according to the chip processing requirements of the SiC substrate. The processing requirements include the target temperature distribution, the allowed temperature gradient range, and the temperature change rate at different stages of the processing process. The power adjustment probability of each heating unit is calculated based on the temperature control target value and the contribution coefficient matrix of the heating unit. The contribution coefficient matrix reflects the influence degree of different heating units on the temperature uniformity of the SiC substrate surface. In the calculation of the power adjustment probability, the contribution coefficient is combined with the temperature target deviation to obtain the power adjustment probability of each heating unit. According to the power adjustment probability, a heating unit power adjustment probability distribution table is constructed, which corresponds the number of each heating unit to its corresponding power adjustment probability. The random selection algorithm is applied to the heating unit power adjustment probability distribution table to determine the set of heating units that need to be adjusted in each control period. The random selection algorithm dynamically selects different heating units for power adjustment according to the set control period, and ensures that the adjustment frequency of different heating units remains balanced in a long processing process. The set of heating units selected in each control period consists of a certain proportion of key heating units and auxiliary heating units, where the key heating units are responsible for precise adjustment of the target hot zone, and the auxiliary heating units are used to maintain overall temperature uniformity. Through the dynamic adjustment of the random selection algorithm, the problem of excessive or insufficient power adjustment frequency of certain heating units in a long-term processing process is effectively avoided, improving the stability and reliability of the heating system. According to the set of heating units that need to be adjusted in each control period, the power adjustment amount calculation is performed on the temperature deviation in the target hot zone distribution diagram to obtain the initial power adjustment matrix. The calculation of the power adjustment amount needs to consider the deviation between the current temperature distribution and the target temperature distribution, and calculate the appropriate power adjustment value according to the size and direction of the temperature deviation. For hot spot areas with large temperature deviation, higher power adjustment amount is allocated to rapidly reduce the temperature deviation, while for areas with small temperature deviation, smaller power adjustment amount is used to maintain temperature uniformity. The initial power adjustment matrix is modified using the coupling compensation coefficient, which is used to quantify the temperature coupling effect between different heating units, thereby preventing temperature fluctuations or local overheating between adjacent heating units during power adjustment. The modified power adjustment matrix is fused and smoothed with the current heating power setting value to obtain the final heating power setting matrix. The fusion and smoothing process effectively suppresses the sudden change phenomenon in the power adjustment process, preventing temperature fluctuations caused by excessive power adjustment amplitude.

[0038] Step S500, four-stage hierarchical control is performed on the heating power setting matrix to obtain a smooth transition control curve.

[0039] Specifically, all heating unit powers in the heating power setting matrix are initialized and set to 70% to 100% of the maximum power, and a reasonable temperature rising rate is set according to the heat capacity, thickness, size and target temperature range of the SiC substrate. This stage is a fast preheating stage, and the goal is to quickly raise the temperature of the SiC substrate from room temperature to the basic temperature interval, while avoiding thermal shock or excessive temperature stress caused by excessive temperature rising rate, which may cause cracking or thermal damage of the SiC substrate. The control parameters of the fast preheating stage include the maximum heating power, the maximum allowed temperature rising rate and the preheating time. By accurately setting these parameters, the SiC substrate can complete the basic temperature rise in a short time. The power distribution in the heating power setting matrix is optimized and adjusted based on the heating unit contribution coefficient matrix to eliminate the temperature unevenness caused by the fast preheating stage. The heating unit contribution coefficient matrix reflects the influence of different heating units on the uniformity of the SiC substrate surface temperature. In the temperature uniformity stabilization stage, the power of the key heating units is adjusted first according to the size of the contribution coefficient, and the power intervention on the non-key area is reduced, thereby improving the uniformity of the overall temperature distribution. The control parameters of the temperature uniformity stabilization stage include the temperature uniformity target value, the heating unit power distribution adjustment proportion and the power adjustment frequency. The optimization of these parameters can effectively reduce the temperature standard deviation of the SiC substrate surface. After the temperature uniformity stabilization stage is completed, the process temperature interval T1 to T2 is segmented, and the temperature interval is divided into M temperature sensitive subintervals and G non-temperature sensitive subintervals. The temperature sensitive subinterval refers to the interval where the temperature change has a significant impact on the processing quality of the SiC substrate, and a lower temperature rising rate is set to ensure the stability of the temperature change, while a higher temperature rising rate is used for the non-temperature sensitive subinterval to speed up the temperature rising speed and shorten the processing time. The control parameters of the accurate temperature rising stage include the number of temperature subintervals, the temperature rising rate of the temperature sensitive subinterval, the temperature rising rate of the non-temperature sensitive subinterval and the temperature rising target value of each subinterval. Through accurate temperature segmentation control, temperature fluctuations in the key temperature interval can be effectively avoided. When the temperature of the SiC substrate reaches the target temperature T2, the process constant temperature stage is entered. The goal of this stage is to maintain the SiC substrate temperature stable within the constant temperature interval required by the processing process, while controlling the temperature fluctuation and temperature gradient within the allowed range. In the process constant temperature stage, the proportional coefficient and the differential coefficient are dynamically adjusted according to the real-time temperature deviation range, so as to realize the adaptive optimization of the temperature control parameters. When the temperature deviation is large, the proportional coefficient is increased to speed up the temperature recovery speed, and when the temperature deviation is small, the proportional coefficient is reduced to avoid excessive temperature fluctuations. At the same time, the differential coefficient is adjusted according to the temperature change trend to enhance the system's ability to suppress temperature fluctuations. The control parameters of the process constant temperature stage include the proportional coefficient, the differential coefficient, the temperature deviation threshold and the adjustment frequency.The control parameters of the rapid preheating stage, the control parameters of the uniform temperature stabilization stage, the control parameters of the precise temperature rising stage and the control parameters of the process constant temperature stage are subjected to data smoothing processing, and combined with time sequence information in the control process, a smooth transition control curve is constructed.

[0040] In step S600, a progressive temperature control strategy is implemented according to the smooth transition control curve, and global temperature control parameters meeting the SiC substrate chip processing precision requirements are obtained.

[0041] Specifically, the surface of the SiC substrate is divided into a core process area, a transition area, and an edge area according to the thermal sensitivity and process requirements, and different areas are assigned corresponding priority weights. The core process area is given the highest priority weight in the temperature control process, the priority weight of the transition area is slightly lower than that of the core process area, and the priority weight of the edge area is the lowest. According to the spatiotemporal evolution control and dynamic allocation control resource strategy of the multi-level temperature control structure, a resource dynamic scheduling matrix is obtained. The spatiotemporal evolution control refers to dynamically adjusting the control strategy of different areas according to the temperature change trend in different processing stages. In the early stage of processing, most of the control resources are concentrated in the core process area to ensure the temperature accuracy of the key area first. In the middle stage of processing, the control range is gradually expanded to the transition area and the edge area to eliminate potential thermal unevenness. In the later stage of processing, the global temperature control resources are evenly distributed to maintain the overall temperature stability of the SiC substrate. The dynamic allocation of control resources automatically adjusts the power regulation frequency and amplitude of the heating unit according to the real-time change of the temperature deviation of each area, and dynamically generates the resource dynamic scheduling matrix. Based on the resource dynamic scheduling matrix, combined with the temperature-power closed-loop response time self-adjusting algorithm, different areas in the multi-level temperature control structure are processed with different control frequencies to obtain a multi-scale temperature field collaborative control network. The temperature-power closed-loop response time self-adjusting algorithm adjusts the power distribution and control frequency according to the dynamic characteristics of the temperature change of different areas. For areas with faster response speed, a lower control frequency is used to reduce unnecessary energy consumption. For areas with slower response speed, the control frequency is increased to ensure that the temperature deviation is effectively suppressed in the shortest time. The multi-scale temperature field collaborative control network realizes the temperature control collaboration between different spatial scale areas, thereby avoiding the occurrence of excessive temperature gradient or local overheating during the heating or cooling process, and improving the overall temperature control accuracy. On the basis of the multi-scale temperature field collaborative control network, a temperature steady-state prediction compensation mechanism is introduced. This mechanism predicts the future temperature trend based on the current temperature distribution state and compensates for potential temperature deviations in advance. The temperature steady-state prediction compensation mechanism forms a two-way regulation system by calculating the weighted combination of temperature deviation, temperature trend, and thermal inertia, and adjusts the power in both directions when the temperature is ahead or lagging, realizing the feedforward and feedback control of temperature change. The two-way regulation system combines the control quality evaluation index with the gradient descent algorithm to perform parameter adaptive optimization, thereby improving the accuracy of global temperature control parameters. The control quality evaluation index includes temperature uniformity, temperature fluctuation amplitude, energy efficiency, and other dimensions. By dynamically evaluating these indicators, the effect of the current temperature control strategy is judged in real time. Combined with the gradient descent algorithm, the temperature control parameters are adjusted according to the trend of the evaluation index to gradually approach the optimal control solution, thereby realizing parameter adaptive optimization. After multiple rounds of iterative optimization, the global temperature control parameters are generated.

[0042] In the embodiment of the present application, by establishing a high-precision three-dimensional temperature field distribution model, real-time temperature prediction is realized, and reliable data support is provided for accurate temperature control in SiC substrate chip processing. Based on a specific temperature parameter analysis process, target parameters such as hot spot position coordinates, average temperature value and maximum temperature gradient can be accurately obtained, and their uncertainty can be quantified, laying a foundation for thermal zone analysis. By numbering and positioning the heating units and applying differentiated weight coefficients, the target heating unit that has a decisive influence on temperature uniformity is accurately identified, forming a heating unit contribution coefficient matrix and a target thermal zone distribution map. Based on this, probability-based power distribution is realized, and the target heating unit is preferentially allocated adjustment resources, while eliminating mutual interference between adjacent heating units to prevent power mutation. A four-stage hierarchical control method is used to realize smooth transition from room temperature to process temperature, without the need for additional preheating equipment, providing a stable processing temperature environment for SiC substrates. The progressive temperature control strategy divides the SiC substrate surface into three control levels, dynamically allocates control resources using a space-time evolution control model, and combines adaptive parameter optimization and a multi-level safety protection system to ensure high precision and high stability of temperature control. In addition, by accurately controlling the temperature distribution of the SiC substrate surface, defects caused by uneven and fluctuating temperature are significantly reduced, improving the yield and consistency of SiC chip manufacturing. At the same time, due to the absence of additional preheating equipment, the optimized power distribution strategy also reduces energy consumption, achieving energy-efficient utilization of the SiC substrate chip processing process.

[0043] In a specific embodiment, the process of step S100 can specifically include the following steps:

[0044] Collecting the thermal physical parameters of the SiC substrate material to obtain a thermal physical property matrix containing thermal conductivity, specific heat capacity and thermal diffusivity;

[0045] Based on the thermal physical property matrix, a set of thermal physical parameters of the SiC substrate at different temperatures is constructed;

[0046] Inputting the set of thermal physical parameters and the geometric structure data of the SiC substrate into a finite element analysis system to obtain an initial three-dimensional model;

[0047] Applying boundary conditions including edge heat dissipation and ambient temperature to the initial three-dimensional model to obtain a temperature field response equation;

[0048] Based on the temperature field response equation, real-time temperature prediction is performed to obtain a surface temperature distribution map and a temperature gradient distribution map of the SiC substrate;

[0049] Comparing and calibrating the surface temperature distribution map and the temperature gradient distribution map with the measured temperature value to obtain a three-dimensional temperature field distribution model.

[0050] Specifically, the thermal physical properties of SiC substrates are measured and data are collected. The thermal physical properties of SiC materials include thermal conductivity, specific heat capacity, and thermal diffusivity, which are key parameters that have a significant impact on the temperature distribution and thermal response of SiC substrates. Thermal conductivity is a parameter that measures the ability of a material to conduct heat, affecting the speed of heat transfer in SiC substrates. Specific heat capacity reflects the ability of a material to raise its temperature after absorbing heat, while thermal diffusivity determines the speed of heat diffusion within the material. Under different temperature conditions, the thermal conductivity of SiC substrates is tested, the specific heat capacity is measured, and the thermal diffusivity is evaluated through steady-state or transient methods, obtaining a thermal physical property matrix covering different temperature ranges. The measurement of thermal conductivity uses the laser flash method or the transient plane heat source method, the specific heat capacity is determined by differential scanning calorimetry, and the thermal diffusivity is obtained by thermal resistance analysis or transient thermal response testing. Through the fitting and interpolation of a large amount of experimental data, a SiC substrate thermal physical property matrix is obtained under different temperature, thickness, and size conditions. Based on the thermal physical property matrix, a set of SiC substrate thermal physical parameters at different temperatures is constructed. Due to the large temperature variation of SiC substrates during processing, their thermal physical property parameters will dynamically change with temperature. The data in the thermal physical property matrix are interpolated to generate a set of thermal physical parameters at different temperature points, which includes the thermal conductivity, specific heat capacity, and thermal diffusivity information of SiC substrates at different temperatures. Through dynamic updating of the thermal physical parameters, the nonlinear variation characteristics of the thermal physical parameters are considered during temperature field simulation, thereby improving the accuracy of the model. The constructed thermal physical parameter set is integrated with the geometric data of the SiC substrate and input into the finite element analysis system to form an initial three-dimensional temperature field model. The geometric data of the SiC substrate includes substrate thickness, size, heater distribution, edge shape, and the relative position of heat sources and heat dissipation areas, which helps to establish an accurate three-dimensional temperature field model. The finite element analysis system generates an initial three-dimensional temperature model containing temperature changes at different locations based on the input thermal physical parameters and geometric data through mesh division and heat flow path calculation. To improve simulation accuracy, non-uniform mesh division is used to divide the hot spot area more meticulously, ensuring that the temperature variation of the hot spot area can be accurately captured. After constructing the initial three-dimensional temperature field model, boundary conditions including edge heat dissipation and ambient temperature are applied to the model to simulate heat conduction and heat dissipation effects in the actual processing environment. The SiC substrate is affected by heat dissipation from the edge during processing, so edge heat dissipation boundary conditions are introduced into the model, and dynamic thermal equilibrium calculations are performed in combination with ambient temperature changes. At the same time, the thermal contact resistance between the heating unit and the SiC substrate is considered. The application of these boundary conditions helps to form a thermodynamic model that is closer to the actual processing environment, improving the accuracy of temperature field simulation. Based on the temperature field model with applied boundary conditions, real-time temperature prediction is achieved by solving the temperature field response equation.The temperature field response equation is a partial differential equation group established based on the heat conduction law and the principle of energy conservation, which can describe the change rule of the internal temperature of the SiC substrate with time and space. By inputting the heating power distribution, the initial temperature field state and the boundary conditions, the temperature field response equation is numerically solved by using a finite element solver to obtain the temperature change of the SiC substrate at different positions and different time points, so as to generate the surface temperature distribution map and the temperature gradient distribution map of the SiC substrate. The surface temperature distribution map reflects the temperature distribution of the SiC substrate under the action of each heating unit, and the temperature gradient distribution map shows the temperature change rate at different positions on the surface of the SiC substrate. The surface temperature distribution map and the temperature gradient distribution map are compared and calibrated with the measured temperature value. The measured temperature data is obtained by a thermocouple or an infrared thermometer, and is compared with the simulated temperature data. The deviation between the measured temperature and the simulated temperature is calculated, and a correction factor is used for dynamic model updating, so as to gradually correct the parameters in the finite element analysis model, so as to ensure that the simulation model can accurately reflect the heat distribution state of the SiC substrate in the real processing process, and finally obtain a three-dimensional temperature field distribution model.

[0051] In a specific embodiment, the process of performing step S200 can specifically include the following steps:

[0052] The heating power distribution matrix and the environmental parameters of the SiC substrate are input into the three-dimensional temperature field distribution model to obtain original temperature distribution data;

[0053] The original temperature distribution data is subjected to feature extraction and noise reduction processing to obtain a temperature distribution feature matrix, and a set of surface hot spot position coordinates of the SiC substrate is calculated based on the temperature distribution feature matrix;

[0054] The set of surface hot spot position coordinates is subjected to temperature gradient analysis to obtain surface temperature change rate data of the SiC substrate;

[0055] The surface temperature change rate data is compared and analyzed with a process temperature threshold to obtain a temperature deviation evaluation result;

[0056] According to the temperature deviation evaluation result, a SiC substrate surface temperature distribution parameter set containing hot spot coordinates, average temperature values, maximum temperature gradients and uncertainty quantization values is generated.

[0057] Specifically, the heating power distribution matrix is preprocessed with environmental parameter data and input into a three-dimensional temperature field distribution model. The heating power distribution matrix is a two-dimensional matrix formed according to the spatial layout of different heating units on the SiC substrate and the power distribution of the heating units, where each matrix element corresponds to the power value of a specific heating unit. The environmental parameters include environmental temperature, heat convection coefficient, edge heat dissipation coefficient, and thermal contact impedance between the heating unit and the SiC substrate, etc. After inputting these data into the three-dimensional temperature field distribution model, the temperature field response equation is solved by the finite element solver to obtain the original temperature distribution data, which includes the temperature change information of the SiC substrate at different spatial positions and different time points, and forms a complete temperature distribution matrix. Feature extraction and noise reduction processing are performed on the original temperature distribution data to obtain a temperature distribution feature matrix reflecting the temperature change characteristics. Since the original temperature distribution data contains a large amount of measurement noise, which comes from factors such as heater power fluctuations, environmental temperature fluctuations, measurement device errors, and uneven thermal contact, a noise reduction algorithm is used to process the data. Noise reduction methods include median filtering, Kalman filtering, and wavelet denoising, etc. Through spatial and temporal filtering of temperature data, noise is effectively removed and the main features of temperature change are preserved. During the noise reduction process, feature extraction is performed to extract key feature components from the original temperature distribution data through principal component analysis or singular value decomposition method, forming a more representative temperature distribution feature matrix to reflect the temperature change trend of the SiC substrate surface in different regions. Based on the temperature distribution feature matrix, the coordinates of the hot spot position set on the SiC substrate surface are calculated. A local maximum value detection algorithm is used to scan the temperature distribution feature matrix to identify the regions with the highest local temperature in the temperature distribution, and the spatial coordinates of these hot spots are extracted to form a coordinate set containing all hot spot positions. The hot spot region is located in the region directly affected by the heating unit or the region with high thermal load, which needs to be focused on in the subsequent power distribution and temperature control process. Temperature gradient analysis is performed on the surface hot spot position coordinate set to calculate the surface temperature change rate data of the SiC substrate. Through temperature gradient analysis of the hot spot region and its adjacent region, the finite difference method or gradient calculation method is used to perform spatial and temporal difference operations on the temperature data of the hot spot region to obtain the temperature change rate data of different hot spot regions, reflecting the thermal change speed of different positions on the SiC substrate surface. The surface temperature change rate data is compared and analyzed with the process temperature threshold to obtain the temperature deviation evaluation result. The process temperature threshold is the temperature change allowed range set according to the SiC substrate processing process requirements, including the maximum temperature gradient, the allowed temperature fluctuation range, and the maximum temperature limit of the hot spot region, etc. By comparing the calculated temperature change rate data with the process temperature threshold point by point, it is determined whether the temperature change of the current SiC substrate surface is within the allowed range, and the temperature deviation amount of different regions is calculated.For the regions exceeding the process temperature threshold, the subsequent power allocation and temperature control strategy is revised to ensure the uniform and stable surface temperature of the SiC substrate during processing. After completing the temperature deviation evaluation, a set of SiC substrate surface temperature distribution parameters including hotspot coordinates, average temperature values, maximum temperature gradients, and their uncertainty quantification values are generated according to the evaluation results. The hotspot coordinates are a set of hotspot positions obtained through the preliminary hotspot identification, accurately marking the region with the highest temperature on the SiC substrate; the average temperature values are obtained by weighted averaging the temperature data in the temperature distribution feature matrix, used to measure the overall temperature level of the SiC substrate surface; the maximum temperature gradient reflects the maximum amplitude of the SiC substrate surface temperature change, occurring at the junction of the hotspot region and the surrounding region; and the uncertainty quantification value is obtained by uncertainty quantification analysis of the temperature data based on the heteroscedastic model or Bayesian estimation method, providing a reference for temperature prediction and adaptive adjustment of the temperature control strategy.

[0058] In a specific embodiment, the process of performing step S300 can specifically include the following steps:

[0059] Numbering and positioning the heating units in the SiC substrate heating system to obtain a heating unit index matrix;

[0060] Calculating the temperature uniformity index based on the hotspot coordinates and average temperature values in the SiC substrate surface temperature distribution parameter set;

[0061] Determining the perturbation range of each heating unit according to the maximum temperature gradient and uncertainty quantification value in the SiC substrate surface temperature distribution parameter set;

[0062] Spatially mapping the hotspot coordinates in the SiC substrate surface temperature distribution parameter set with the heating unit index matrix to obtain a target heating unit set;

[0063] Setting differentiated position weight factors for the target heating unit set and non-target region heating units according to the temperature uniformity index and perturbation range to obtain a heating unit contribution coefficient matrix;

[0064] Performing heat zone contribution thermal map analysis according to the heating unit contribution coefficient matrix, while labeling the hotspot coordinates in the SiC substrate surface temperature distribution parameter set to obtain a target heat zone distribution map.

[0065] Specifically, all heating units in the heating system are accurately numbered and positioned. The heating units are uniformly arranged in a matrix form under the SiC substrate, and the power of each heating unit is independently controlled. Therefore, a heating unit index matrix is constructed to identify all heating units in a numbered manner and record their coordinate positions in space. The dimensions of the heating unit index matrix are determined by the number of heating units in the x and y directions, where each element in the matrix corresponds to the heating unit number at a specific location, along with its coordinate information. By establishing the heating unit index matrix, a precise reference framework is provided for subsequent heating unit positioning, temperature change analysis, and power distribution. Based on the hotspot coordinates and average temperature values in the temperature distribution parameter set of the SiC substrate surface, a temperature uniformity index is calculated to quantify the uniformity of the current temperature distribution. The temperature uniformity index uses the temperature standard deviation or coefficient of variation to measure, by calculating the deviation between the temperature at each position on the SiC substrate surface and the average temperature, a value reflecting the uniformity of the temperature distribution. The closer the value of the temperature uniformity index to zero, the more uniform the temperature distribution, while a larger value indicates that there is a large temperature gradient or hotspot region on the SiC substrate surface, requiring further adjustment of the heating power. While calculating the temperature uniformity index, the hotspot coordinates on the SiC substrate surface are identified, and combined with the temperature change characteristics of the hotspot region, data support is provided for subsequent heat zone contribution analysis. According to the maximum temperature gradient and uncertainty quantification value in the temperature distribution parameter set of the SiC substrate surface, the disturbance range of each heating unit is determined. The maximum temperature gradient reflects the temperature variation amplitude between different regions on the SiC substrate surface, while the uncertainty quantification value is used to evaluate the error range of temperature measurement or model prediction. By combining the maximum temperature gradient with the uncertainty quantification value, a disturbance range for each heating unit is set, which represents the amount of power adjustment that the heating unit can withstand during actual operation, to ensure that temperature jumps or system instability are not caused during power adjustment. The disturbance range of the heating unit is obtained based on multiple disturbance experiments and temperature response simulations. By performing ±5%, ±10%, and ±15% power disturbances on different heating units and analyzing their impact on the temperature distribution of the SiC substrate surface, a data set containing the power adjustment limits of each heating unit is formed. The hotspot coordinates in the temperature distribution parameter set of the SiC substrate surface are spatially mapped to the heating unit index matrix to obtain the target heating unit set. Spatial mapping is achieved by calculating the spatial distance relationship between the hotspot region and the heating units, and the Euclidean distance calculation method is used to map the hotspot coordinates to the nearest heating unit number. The target heating unit set is a collection of heating units that directly affect the temperature change of the hotspot region, and the power adjustment of these heating units is crucial for maintaining the temperature stability of the hotspot region, which needs to be considered in the subsequent heat zone contribution analysis.According to the temperature uniformity index and the disturbance range, a differentiated position weight factor is set for the target heating unit set and the non-target area heating unit to obtain a heating unit contribution degree coefficient matrix. The position weight factor is used to reflect the influence degree of different regional heating units on temperature distribution. For the units in the target heating unit set, higher weight factors are given because of their strong control effect on the hot spot area. For the heating units in the non-target area, lower weight factors are set because of their weak influence on the overall temperature change. By differentiating the weight distribution of the target heating unit and the non-target area heating unit, the distribution of heating power is optimized, thereby improving the overall efficiency of the temperature control system. Based on the heating unit contribution degree coefficient matrix, a heat zone contribution degree heat map analysis is performed, and the hot spot coordinates in the SiC substrate surface temperature distribution parameter set are marked in the heat map to obtain a target heat zone distribution map. The heat zone contribution degree heat map is an image generated by two-dimensional interpolation processing of the data in the heating unit contribution degree coefficient matrix, which shows the influence degree of different heating units on temperature distribution and highlights the position and temperature change characteristics of the hot spot area.

[0066] In a specific embodiment, the process of performing step S400 can specifically include the following steps:

[0067] According to the chip processing process requirements of the SiC substrate, a temperature control target value is determined;

[0068] Based on the temperature control target value and the heating unit contribution degree coefficient matrix, the power adjustment probability of each heating unit is calculated, and a heating unit power adjustment probability distribution table is constructed according to the power adjustment probability;

[0069] A random selection algorithm is applied to the heating unit power adjustment probability distribution table to obtain a heating unit set that needs to be adjusted in each control cycle;

[0070] According to the heating unit set that needs to be adjusted in each control cycle, a power adjustment amount calculation is performed on the temperature deviation in the target heat zone distribution map to obtain an initial power adjustment matrix;

[0071] A coupling compensation coefficient is used to modify the initial power adjustment matrix to obtain a modified power adjustment matrix, and the modified power adjustment matrix is fused and smoothed with the current heating power set value to obtain a heating power set matrix.

[0072] Specifically, the temperature control target value is determined according to the chip processing requirements of the SiC substrate, including the target temperature range required during processing, the maximum temperature gradient allowed, the temperature upper limit of the hot spot area, and the maximum tolerance range of temperature fluctuations. Since the chip processing of the SiC substrate is carried out in a high-temperature environment, different processing techniques have different requirements for temperature uniformity, heating rate, and steady-state temperature accuracy. Therefore, the temperature control target value is accurately set according to the specific process requirements. The setting of the temperature target value should fully consider the thickness of the SiC substrate, the thermal properties of the material, the layout of the heating unit, and the environmental thermal load, so as to ensure that the target temperature distribution can meet the strict requirements for thermal uniformity and processing accuracy during chip processing. Based on the temperature control target value and the heating unit contribution coefficient matrix, the power adjustment probability of each heating unit is calculated. The heating unit contribution coefficient matrix reflects the influence degree of different heating units on the temperature distribution of the SiC substrate surface. According to the size of the contribution coefficient, each heating unit is assigned a corresponding power adjustment probability. The higher the contribution of the heating unit to temperature change, the greater the power adjustment probability, so the higher the possibility of these heating units being selected for power adjustment in each control cycle. In the process of calculating the power adjustment probability, the heating unit contribution coefficient matrix is combined with the temperature control target value to obtain a probability matrix containing the power adjustment probabilities of all heating units through weighted calculation, forming a power adjustment probability distribution table. The random selection algorithm is applied to the power adjustment probability distribution table of the heating unit to determine the heating unit set that needs to be adjusted in each control cycle. The introduction of the random selection algorithm is to realize dynamic adjustment in the temperature control process, so that the system can adaptively adjust the power distribution of the heating unit according to the real-time temperature change, thereby avoiding the problem of local overheating or temperature control accuracy reduction caused by fixed heating unit adjustment. The random selection algorithm randomly selects a part of the heating units for power adjustment according to the weight of the power adjustment probability distribution table in each control cycle. This selected heating unit set includes key heating units in the hot spot area and auxiliary heating units in part of the non-critical area, to achieve balanced control of the overall temperature field. The random selection algorithm effectively improves the robustness of the system and ensures the workload balance of the heating units during long-term operation, thereby improving the stability of the temperature distribution on the surface of the SiC substrate. After determining the heating unit set that needs to be adjusted in each control cycle, the power adjustment amount calculation is performed on the temperature deviation in the target thermal zone distribution according to the heating unit set, to obtain the initial power adjustment matrix. The calculation of the power adjustment amount needs to combine the deviation between the current temperature distribution and the target temperature distribution, and dynamically adjust according to the size, direction, and spatial distribution of the temperature deviation. For the hot spot area, a high power adjustment amount is allocated to quickly reduce the temperature of the hot spot area, while for the non-hot spot area, a small power adjustment amount is allocated to maintain the uniformity of the overall temperature distribution.The initial power adjustment matrix determines the power adjustment amount of each heating unit by analyzing the gradient change of the temperature distribution, forming a two-dimensional matrix reflecting the power adjustment amplitude and direction. Considering the thermal coupling effect between heating units, a coupling compensation coefficient is used to modify the initial power adjustment matrix. There is mutual thermal coupling effect between heating units, and the power adjustment of one heating unit will have a chain reaction on the temperature distribution of adjacent units. Therefore, in the power adjustment process, the initial power adjustment matrix is modified by introducing a coupling compensation coefficient to avoid temperature fluctuations or unevenness caused by coupling effect during power adjustment. The coupling compensation coefficient is obtained by modeling and simulating the heat conduction relationship between heating units, reflecting the thermal interaction between different heating units and dynamically compensating the thermal influence between heating units during power adjustment. The modified power adjustment matrix is more consistent with the actual temperature change, thereby improving the accuracy and stability of power adjustment. The modified power adjustment matrix is fused and smoothed with the current heating power setting value. The modified power adjustment matrix is smoothed by time-weighted average, moving average or exponential smoothing, etc. to ensure smooth and stable power change process and avoid system instability caused by sudden power change. The final heating power setting matrix is used for temperature control in the next control period.

[0073] In this embodiment, before the four-stage hierarchical control of the heating power setting matrix based on the heating unit contribution degree coefficient matrix and the target heat zone distribution map is calculated, the following steps are included: applying the progressive polyhedral approximation method to the optimized heating power setting matrix to obtain a linearized temperature control problem expression; dividing the SiC substrate surface into multiple temperature control grid cells, assigning an approximation accuracy parameter based on the temperature control difficulty of each grid cell to obtain an adaptive grid density distribution map; according to the adaptive grid density distribution map, using high-density piecewise linear approximation in key heat zones and low-density piecewise linear approximation in non-key areas to obtain a position-adaptive piecewise linear constraint set; inputting the piecewise linear constraint set into the optimization solver to perform fast solving calculation to obtain an initial linear power control strategy; applying the initial linear power control strategy to the SiC substrate for preliminary temperature adjustment and recording the deviation value between the actual temperature distribution and the predicted temperature distribution to obtain a temperature control error distribution map; identifying the control area that needs to be refined according to the temperature control error distribution map, refining the piecewise linear constraint in the area where the error exceeds the preset threshold to obtain an iteratively optimized refined linear constraint set; applying the iteratively optimized refined linear constraint set to the subsequent control period to gradually converge to a near-optimal solution through multiple iterations to obtain a high-precision linear temperature control strategy; adjusting the optimized heating power setting matrix based on the high-precision linear temperature control strategy to obtain the final heating power setting matrix as the input data for the four-stage hierarchical control.

[0074] In a specific embodiment, the process of performing step S500 can specifically include the following steps:

[0075] Set the maximum power and the temperature rising rate for all heating unit powers in the heating power setting matrix simultaneously to obtain the control parameters of the rapid preheating stage;

[0076] Adjust the power distribution in the heating power setting matrix based on the heating unit contribution degree coefficient matrix to obtain the control parameters of the uniform temperature stabilization stage;

[0077] Divide the process temperature interval T1 to T2 into M temperature sensitive subintervals and G non-temperature sensitive subintervals, and set the first temperature rising rate of the temperature sensitive subintervals and the second temperature rising rate of the non-temperature sensitive subintervals to obtain the control parameters of the accurate temperature rising stage;

[0078] Dynamically adjust the proportional coefficient and the differential coefficient according to the temperature deviation range to obtain the control parameters of the process constant temperature stage;

[0079] Perform data smoothing processing and time sequence curve construction on the control parameters of the rapid preheating stage, the control parameters of the uniform temperature stabilization stage, the control parameters of the accurate temperature rising stage, and the control parameters of the process constant temperature stage to obtain a smooth transition control curve.

[0080] Specifically, according to the material properties, thickness, size of the SiC substrate and the specific requirements of the processing technology, the target temperature and the temperature rising time of the rapid preheating stage are determined. In the rapid preheating stage, the temperature of the SiC substrate is rapidly raised from room temperature to a stable basic temperature interval T1 between 100°C and 150°C. This stage is achieved by setting all heating unit power to 70% to 100% of the maximum power at the same time. At the same time, in order to prevent thermal shock from causing thermal stress or cracks in the SiC substrate, the maximum temperature rising rate is limited, and the temperature rising rate is set to 40°C / min to 50°C / min to ensure the structural stability of the SiC substrate during temperature rising. Through the rapid heating strategy, the temperature of the SiC substrate is raised to the initial temperature required for the uniform stable stage in a short time. After completing the rapid preheating stage, enter the uniform stable stage, and optimize and adjust the power distribution in the heating power setting matrix based on the contribution coefficient matrix of the heating unit. The contribution coefficient matrix of the heating unit reflects the degree of influence of different heating units on the uniformity of the surface temperature of the SiC substrate. In the uniform stable stage, the power of the key heating unit is adjusted first according to the size of the contribution coefficient, and the power distribution of the non-key area is reduced to realize the uniformization of the surface temperature of the SiC substrate. The goal of the uniform stable stage is to reduce the standard deviation of the surface temperature of the SiC substrate to the set threshold range, requiring that the temperature standard deviation does not exceed 1.5°C. In this stage, by gradually reducing the power of the hot spot area heating unit, while appropriately increasing the power of the edge area and the lower temperature area, the non-uniformity in the temperature distribution is eliminated, and the goal of temperature uniformization is achieved. The control parameters of the uniform stable stage include the power distribution ratio of the heating unit, the temperature uniformity evaluation standard and the power adjustment frequency, and reasonable setting of these parameters can significantly improve the precision of temperature control. After entering the precise temperature rising stage, the process temperature interval T1 to T2 is divided into M temperature sensitive subintervals and G non-temperature sensitive subintervals, and different temperature rising rates are set to meet the precise control requirements of the temperature change of the SiC substrate in the processing process. The temperature sensitive subinterval refers to the temperature interval in which the temperature change has a significant impact on the performance, structural stability and processing quality of the SiC substrate. In these intervals, a lower temperature rising rate is used to avoid thermal stress or thermal deformation of the SiC substrate caused by rapid temperature change. In the temperature sensitive subinterval, the temperature rising rate is set to 10°C / min to 20°C / min, while in the non-temperature sensitive subinterval, since the temperature change has less impact on the processing quality, a higher temperature rising rate is used, which is 30°C / min to 40°C / min. By precisely dividing the process temperature interval into sensitive and non-sensitive intervals and setting different temperature rising rates, the temperature change of the SiC substrate in the key temperature interval is precisely controlled, and the temperature rising speed is accelerated in the non-key interval, thereby shortening the overall temperature rising time and improving the processing efficiency.The control parameters of the precise temperature rising stage include the number of subintervals, the temperature rising target value of each subinterval, the temperature rising rate of the temperature sensitive subinterval, and the temperature rising rate of the non-temperature sensitive subinterval. The precise setting of these parameters can effectively improve the response speed and accuracy of the temperature control system, and ensure that the temperature change of the SiC substrate during the temperature rising process meets the process requirements. After the temperature of the SiC substrate reaches the target temperature T2, the process constant temperature stage is entered. The goal of this stage is to maintain the temperature of the SiC substrate stable within the target temperature range, and control the temperature fluctuation and temperature gradient within the allowable range. In the process constant temperature stage, the proportional coefficient and the differential coefficient are dynamically adjusted according to the temperature deviation range, to realize the adaptive optimization of the temperature control parameters. The role of the proportional coefficient is to adjust the power output of the heating unit according to the size of the temperature deviation. When the temperature deviation is large, the proportional coefficient is increased to speed up the temperature recovery speed. When the temperature deviation is small, the proportional coefficient should be reduced to prevent the power adjustment amplitude from being too large to cause temperature fluctuation. The role of the differential coefficient is to suppress the temperature change trend. Through dynamic analysis of the temperature change rate, when the temperature change speed is fast, the differential coefficient is increased to suppress the temperature fluctuation. When the temperature change trend tends to be stable, the differential coefficient is reduced to maintain the temperature stability. The control parameters of the process constant temperature stage include the proportional coefficient, the differential coefficient, the temperature deviation threshold, and the adjustment frequency. By dynamically adjusting these parameters, precise control of the temperature of the SiC substrate is realized, ensuring that the temperature fluctuation amplitude does not exceed ±0.5℃, and maintaining the temperature stability during the entire processing process. The control parameters of the rapid preheating stage, the control parameters of the uniform temperature stabilization stage, the control parameters of the precise temperature rising stage, and the control parameters of the process constant temperature stage are subjected to data smoothing processing, and combined with the time sequence information in the control process, a smooth transition control curve is constructed. The data smoothing processing adopts time weighted average, moving average or exponential smoothing algorithm. By smoothing the control parameters of different stages, the sudden change phenomenon caused by rapid parameter change or large power adjustment amplitude in the control process is effectively eliminated, forming a smooth and continuous temperature change curve. At the same time, the time sequence curve is constructed based on the control time nodes of each stage, and the control parameters of each stage are connected in time sequence to form a complete temperature control curve, providing an accurate and stable temperature control trajectory for the SiC substrate chip processing process.

[0081] In a specific embodiment, the process of step S600 can specifically include the following steps:

[0082] According to the heat sensitivity and the processing process requirements, the surface of the SiC substrate is divided into a core process area, a transition area and an edge area, and priority weights are assigned to the core process area, the transition area and the edge area, to obtain a multi-level temperature control structure;

[0083] According to the multi-level temperature control structure, time-space evolution control and dynamic allocation control resources are performed to obtain a resource dynamic scheduling matrix;

[0084] Based on the resource dynamic scheduling matrix and the temperature-power closed-loop response time self-adjusting algorithm, different regions in the multi-level temperature control structure are processed differently in frequency to obtain a multi-scale temperature field cooperative control network;

[0085] A temperature steady-state prediction compensation mechanism is introduced for the multi-scale temperature field cooperative control network, and a weighted combination of temperature deviation, temperature trend and thermal inertia is calculated to obtain a bidirectional regulation system.

[0086] The bidirectional regulation system is used to combine the control quality evaluation index with the gradient descent algorithm to perform parameter adaptive optimization, and obtain global temperature control parameters that meet the SiC substrate chip processing precision requirements.

[0087] Specifically, the thermal sensitivity distribution of the SiC substrate is analyzed, and combined with the specific requirements of the processing technology, different regions are accurately divided. The core process area is the most critical area of chip function, and the temperature change of this area has a direct impact on the processing quality and chip performance, so it is given the highest priority weight to ensure that the temperature of this area remains in the optimal range during temperature control. The transition area is the intermediate area connecting the core process area and the edge area, and its temperature change will not directly affect the chip processing quality, but due to its heat diffusion characteristics, it will indirectly affect the thermal environment of the core process area, so it is assigned a medium priority weight to maintain better temperature uniformity during processing. The edge area is the most significant area of heat dissipation, and due to the rapid loss of heat through the edge, the temperature change of this area has less impact on overall temperature control, so it is assigned the lowest priority weight. The division of the multi-level temperature control structure is automatically adjusted according to the thermal physical properties of the SiC substrate and the processing technology parameters, ensuring that the temperature control accuracy of different regions matches the thermal sensitivity requirements. According to the spatiotemporal evolution control of the multi-level temperature control structure and the dynamic allocation of control resources, a resource dynamic scheduling matrix is obtained. Spatiotemporal evolution control refers to adjusting the control strategy of different regions gradually according to the dynamic changes of SiC substrate temperature distribution at different processing stages. In the early stages of processing, due to the large temperature change, most of the control resources are concentrated in the core process area to prioritize temperature accuracy in the critical area. In the middle stages of processing, as the temperature distribution gradually becomes uniform, the allocation range of control resources gradually expands to the transition area to further optimize the temperature distribution. In the later stages of processing, some control resources are allocated to the edge area to maintain the thermal balance of the overall temperature. The process of dynamically allocating control resources is achieved by analyzing the temperature change rate, power response time, and temperature deviation amplitude of each region, which generates a resource dynamic scheduling matrix. This matrix dynamically adjusts the allocation proportion of control resources at different time nodes, ensuring that different regions receive optimal control resource allocation during processing. Based on the resource dynamic scheduling matrix and the temperature-power closed-loop response time self-adjustment algorithm, different regions in the multi-level temperature control structure are controlled with different frequencies to form a multi-scale temperature field collaborative control network. The temperature-power closed-loop response time self-adjustment algorithm adaptively adjusts the response time and control frequency of power allocation according to the thermal inertia and temperature change characteristics of different regions. In regions with large thermal inertia (such as the core process area), the temperature changes relatively slowly, so a lower control frequency is used to avoid system oscillation caused by frequent power adjustments. In regions with small thermal inertia (such as the edge area), the temperature changes quickly, so a higher control frequency is used to respond to temperature changes and adjust power in a timely manner. The multi-scale temperature field collaborative control network achieves collaborative optimization of temperature control between different regions through differentiated control frequency processing, thereby improving the stability and accuracy of temperature control.On the basis of the multi-scale temperature field cooperative control network, a temperature steady-state prediction compensation mechanism is introduced. The mechanism predicts the future temperature variation trend based on the current temperature distribution data and compensates for the potential temperature deviation in advance. The temperature steady-state prediction compensation mechanism forms a bidirectional regulation system by calculating the weighted combination of temperature deviation, temperature trend and thermal inertia, adjusts the power in both directions when the temperature is ahead or lagging, and realizes the feedforward and feedback control of temperature variation. Specifically, the temperature deviation refers to the real-time difference between the current temperature and the target temperature, the temperature trend is the dynamic analysis result of the current temperature variation rate, and the thermal inertia reflects the thermal response characteristics of different regions. After the three are weighted and combined, a more accurate temperature variation prediction model is formed, and the heating power is adjusted in advance to prevent temperature fluctuations from exceeding the allowed range. The bidirectional regulation system can quickly correct the temperature deviation and suppress potential temperature fluctuations in advance through the predictive compensation mechanism, thereby improving the stability of the temperature control system. After implementing the temperature steady-state prediction compensation, the bidirectional regulation system combines the control quality evaluation index with the gradient descent algorithm to perform parameter adaptive optimization, and obtains the global temperature control parameters that meet the SiC substrate chip processing precision requirements. The control quality evaluation index includes temperature uniformity, temperature fluctuation amplitude, energy efficiency and power adjustment frequency, and other multiple dimensions. These indicators dynamically reflect the overall control effect of the temperature control system. Combined with the gradient descent algorithm, the control parameters are iteratively optimized according to the change trend of the control quality evaluation index. By continuously adjusting the proportion coefficient, the differential coefficient, the response time and the control frequency and other key parameters, the system gradually approaches the optimal control solution. The gradient descent algorithm calculates the gradient information of the control quality evaluation index to determine the current parameter adjustment direction and dynamically correct the control strategy, thereby realizing parameter adaptive optimization. After multiple rounds of iterative optimization, the finally generated global temperature control parameters ensure the uniformity of the temperature distribution, the stability of the temperature fluctuation and the optimization of the energy efficiency during the SiC substrate chip processing, thereby improving the yield and consistency of the SiC substrate chip processing.

[0088] Please refer to Figure 2 , Figure 2 The structural schematic block diagram of the chip processing platform 200 based on intelligent temperature control heating of SiC substrate provided by the embodiment of the present application is shown in Figure 2 As shown in the figure, the chip processing platform 200 based on intelligent temperature control heating of SiC substrate includes:

[0089] The modeling module 201 is configured to model the temperature distribution characteristics of the SiC substrate to obtain a three-dimensional temperature field distribution model.

[0090] The temperature parameter analysis module 202 is configured to perform temperature parameter analysis based on the three-dimensional temperature field distribution model to obtain a set of SiC substrate surface temperature distribution parameters.

[0091] The thermal zone analysis module 203 is configured to perform thermal zone analysis on the SiC substrate surface temperature distribution parameter set to obtain a heating unit contribution degree coefficient matrix and a target thermal zone distribution map.

[0092] The calculation module 204 is configured to calculate a power distribution value based on the heating unit contribution degree coefficient matrix and the target thermal zone distribution map to obtain a heating power setting matrix.

[0093] The hierarchical control module 205 is configured to perform four-stage hierarchical control on the heating power setting matrix to obtain a smooth transition control curve.

[0094] The temperature control module 206 is configured to implement a progressive temperature control strategy according to the smooth transition control curve to obtain global temperature control parameters that meet the SiC substrate chip processing precision requirements.

[0095] Through the cooperation of the above-mentioned components, by establishing a high-precision three-dimensional temperature field distribution model, real-time temperature prediction is realized, and reliable data support is provided for accurate temperature control in SiC substrate chip processing. Based on a specific temperature parameter analysis process, target parameters such as hot spot position coordinates, average temperature values, and maximum temperature gradients can be accurately obtained, and their uncertainties can be quantified, laying a foundation for thermal zone analysis. By numbering and positioning the heating units and applying differentiated weight coefficients, the target heating units that have a decisive influence on temperature uniformity are accurately identified, forming a heating unit contribution degree coefficient matrix and a target thermal zone distribution map. Based on this, a probability-based power distribution is realized, which preferentially allocates adjustment resources to target heating units while eliminating mutual interference between adjacent heating units to prevent power surges. A four-stage hierarchical control method is used to achieve smooth transition from room temperature to process temperature, eliminating the need for additional preheating equipment and providing a stable processing temperature environment for SiC substrates. The progressive temperature control strategy divides the SiC substrate surface into three control levels, uses a spatiotemporal evolution control model to dynamically allocate control resources, combines adaptive parameter optimization and a multi-level safety guarantee system to ensure high precision and high stability of temperature control. In addition, by accurately controlling the SiC substrate surface temperature distribution, defects caused by temperature non-uniformity and fluctuations are significantly reduced, improving SiC chip manufacturing yield and consistency. At the same time, due to the elimination of the need for additional preheating equipment, the optimized power distribution strategy also reduces energy consumption, achieving energy-efficient utilization of the SiC substrate chip processing process.

[0096] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described system, system and unit can refer to the corresponding processes in the foregoing method embodiments, which will not be described here.

[0097] The integrated unit, if implemented in the form of a software function unit and sold or used as an independent product, can be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the present application or the entire or part of the technical solutions that essentially contribute to the prior art can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in the various embodiments of the present application. The aforementioned storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and various media that can store program codes.

[0098] The above description and the above embodiments are only used to illustrate the technical solutions of the present application, but not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacements for part of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for processing a chip based on SiC substrate with intelligent temperature control heating, characterized in that, The method comprises the following steps: Modeling the temperature distribution characteristics of the SiC substrate to obtain a three-dimensional temperature field distribution model; Performing temperature parameter analysis based on the three-dimensional temperature field distribution model to obtain a set of SiC substrate surface temperature distribution parameters; Performing thermal zone analysis on the set of SiC substrate surface temperature distribution parameters to obtain a heating unit contribution degree coefficient matrix and a target thermal zone distribution map; specifically including: calculating a temperature uniformity index and a perturbation range based on the set of SiC substrate surface temperature distribution parameters; creating a heating unit contribution degree coefficient matrix according to the temperature uniformity index and the perturbation range; performing thermal zone contribution degree thermal map analysis according to the heating unit contribution degree coefficient matrix to obtain a target thermal zone distribution map; Calculating a power distribution value based on the heating unit contribution degree coefficient matrix and the target thermal zone distribution map to obtain a heating power setting matrix; specifically including: calculating a set of heating units that need to be adjusted in each control period based on a temperature control target value and the heating unit contribution degree coefficient matrix; performing power adjustment amount calculation on the temperature deviation in the target thermal zone distribution map according to the set of heating units that need to be adjusted in each control period to obtain a heating power setting matrix; Performing four-stage hierarchical control on the heating power setting matrix to obtain a smooth transition control curve; specifically including: setting the maximum power and the temperature rising rate of all heating units in the heating power setting matrix at the same time to obtain the control parameters of the rapid preheating stage; adjusting the power distribution in the heating power setting matrix based on the heating unit contribution degree coefficient matrix to obtain the control parameters of the uniform temperature stabilization stage; dividing the process temperature interval T1 to T2 into M temperature sensitive subintervals and G non-temperature sensitive subintervals, and setting the first temperature rising rate of the temperature sensitive subintervals and the second temperature rising rate of the non-temperature sensitive subintervals to obtain the control parameters of the accurate temperature rising stage; dynamically adjusting the proportional coefficient and the differential coefficient according to the temperature deviation range to obtain the control parameters of the process constant temperature stage; performing data smoothing processing and time sequence curve construction on the control parameters of the rapid preheating stage, the control parameters of the uniform temperature stabilization stage, the control parameters of the accurate temperature rising stage, and the control parameters of the process constant temperature stage to obtain a smooth transition control curve; According to the smooth transition control curve, a progressive temperature control strategy is implemented to obtain global temperature control parameters that meet the SiC substrate chip processing precision requirements; specifically including: according to the thermal sensitivity and processing technology requirements, the surface of the SiC substrate is divided into a core process area, a transition area and an edge area, and priority weights are assigned to the core process area, the transition area and the edge area to obtain a multi-level temperature control structure; according to the multi-level temperature control structure, space-time evolution control and dynamic allocation of control resources are performed to obtain a resource dynamic scheduling matrix; based on the resource dynamic scheduling matrix and a temperature-power closed-loop response time self-adjusting algorithm, different regions in the multi-level temperature control structure are subjected to differentiated control frequency processing to obtain a multi-scale temperature field cooperative control network; a temperature steady-state prediction compensation mechanism is introduced for the multi-scale temperature field cooperative control network, and a weighted combination of temperature deviation, temperature trend and thermal inertia is calculated to obtain a bidirectional regulation system; the bidirectional regulation system is used to combine control quality evaluation indicators with gradient descent algorithm to perform parameter adaptive optimization to obtain global temperature control parameters that meet the SiC substrate chip processing precision requirements.

2. The SiC substrate-based smart temperature-controlled heating chip processing method according to claim 1, wherein, The SiC substrate temperature distribution characteristics are modeled to obtain a three-dimensional temperature field distribution model, including: The thermal physical parameters of the SiC substrate material are collected to obtain a thermal physical property matrix including thermal conductivity, specific heat capacity and thermal diffusivity; Based on the thermal physical property matrix, a set of thermal physical parameters of the SiC substrate at different temperatures is constructed; The set of thermal physical parameters and the geometric structure data of the SiC substrate are input into a finite element analysis system to obtain an initial three-dimensional model; Boundary conditions including edge heat dissipation and ambient temperature are applied to the initial three-dimensional model to obtain a temperature field response equation; Based on the temperature field response equation, real-time temperature prediction is performed to obtain a surface temperature distribution map and a temperature gradient distribution map of the SiC substrate; The surface temperature distribution map and the temperature gradient distribution map are compared and calibrated with the measured temperature values to obtain a three-dimensional temperature field distribution model.

3. The SiC substrate based smart temperature controlled heating chip processing method of claim 1, wherein, Based on the three-dimensional temperature field distribution model, temperature parameter analysis is performed to obtain a SiC substrate surface temperature distribution parameter set, including: The heating power distribution matrix of the SiC substrate and the environmental parameters are input into the three-dimensional temperature field distribution model to obtain original temperature distribution data; The original temperature distribution data are subjected to feature extraction and noise reduction processing to obtain a temperature distribution feature matrix, and based on the temperature distribution feature matrix, a set of surface hot spot position coordinates of the SiC substrate is calculated; Temperature gradient analysis is performed on the set of surface hot spot position coordinates to obtain surface temperature change rate data of the SiC substrate; The surface temperature change rate data are compared and analyzed with the process temperature threshold to obtain a temperature deviation evaluation result; According to the temperature deviation evaluation result, a SiC substrate surface temperature distribution parameter set containing hot spot coordinates, average temperature values, maximum temperature gradients and their uncertainty quantification values is generated.

4. A SiC substrate based smart temperature controlled heating chip processing platform, characterized in that, A chip processing method for performing the SiC substrate based smart temperature control heating as claimed in any one of claims 1-3, the SiC substrate based smart temperature control heating chip processing platform comprising: a modeling module for modeling SiC substrate temperature distribution characteristics to obtain a three-dimensional temperature field distribution model; a temperature parameter analysis module for performing temperature parameter analysis based on the three-dimensional temperature field distribution model to obtain a set of SiC substrate surface temperature distribution parameters; a thermal zone analysis module for performing thermal zone analysis on the set of SiC substrate surface temperature distribution parameters to obtain a heating unit contribution degree coefficient matrix and a target thermal zone distribution map; a calculation module for calculating power allocation values based on the heating unit contribution degree coefficient matrix and the target thermal zone distribution map to obtain a heating power setting matrix; a hierarchical control module for performing four-stage hierarchical control on the heating power setting matrix to obtain a smooth transition control curve; a temperature control module for implementing a progressive temperature control strategy according to the smooth transition control curve to obtain global temperature control parameters that meet SiC substrate chip processing precision requirements.

Images