PID (Proportion Integration Differentiation) temperature control system and method for semiconductor etching machine

Abstract

The invention discloses a PID temperature control system and method for a semiconductor etching machine, and relates to the field of temperature control, and the system comprises a coupling relation construction module, a temperature sensing confidence module, a decoupling compensation module, a coupling relation dynamic updating module, and a transient temperature change control module. According to the invention, by constructing the coupling relation matrix and the inverse matrix thereof, decoupling control of multi-area temperature control is realized, cross interference is effectively suppressed, and the temperature control precision is improved. And a multi-point temperature measurement fusion and anti-interference algorithm is introduced, so that the temperature measurement accuracy and the system robustness are enhanced. A response residual vector is associated with a control vector, and a coupling matrix is dynamically optimized, so that the system has self-learning and model self-adaption capabilities and adapts to working condition changes. And a transient temperature change quick response mechanism is set, so that the control priority can be improved and the PID parameters can be adjusted when the temperature changes suddenly, the response speed and the control stability of the system to abnormal temperature change are remarkably improved, and the thermal uniformity and the process consistency in the etching process are ensured.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of temperature control, in particular to a PID temperature control system and method for a semiconductor etching machine. BACKGROUND

[0002] In the semiconductor manufacturing process, etching process as one of the key steps, puts forward very high precision, stability and response speed requirements to the temperature control in the process cavity. The etching machine cavity is often provided with multiple independent heating areas to ensure the spatial uniformity of the surface temperature of different areas, and then ensure the process repeatability and etching morphology consistency. The temperature control system usually uses multiple heating units to cooperate with their respective PID (Proportional-Integral-Derivative) controllers for adjustment, the goal is to make the actual temperature of each temperature control area stable and track the set target temperature.

[0003] Prior art such as patent for invention with publication number CN119472867B, is a high-temperature superconducting magnet temperature control system based on PID, relates to the technical field of temperature control, comprising: a data acquisition module: for acquiring real-time temperature data of the target high-temperature superconducting magnet, obtaining a first control temperature; an algorithm optimization module: for obtaining a corresponding initial PID control algorithm based on the magnet characteristics of the target high-temperature superconducting magnet, and training the initial PID control algorithm to obtain a first PID control algorithm; a temperature control module: for comparing the first control temperature with a preset target temperature, and inputting the first control temperature into the first PID control algorithm based on the comparison result, and performing temperature control; a state warning module: for monitoring the real-time running state of the target high-temperature superconducting magnet, so as to determine the stability of the target high-temperature superconducting magnet, and perform state display or state warning.

[0004] Prior art such as patent for invention with publication number CN113760007B, is a PID temperature control method, a PID temperature control system and a wavelength selection switch, for controlling the temperature of the center of the cavity of the wavelength selection switch, comprising: a temperature sensor measures the temperature of the heating sheet and transmits the measured heating sheet temperature to the controller; the controller receives the heating sheet temperature and calculates the real-time temperature of the center of the cavity of the wavelength selection switch through the heating sheet temperature, and the controller performs proportional adjustment and integral adjustment on the real-time temperature of the center of the cavity of the wavelength selection switch and then outputs a control signal to the actuator, wherein the controller separates the temperature of the center of the cavity of the wavelength selection switch before performing integral adjustment on the real-time temperature of the center of the cavity of the wavelength selection switch; the actuator controls the temperature of the center of the cavity of the wavelength selection switch to reach the target temperature according to the control signal.

[0005] Based on the above, it can be seen that the prior art in the field of PID temperature control is usually equipped with state monitoring and early warning mechanism to ensure temperature stability, but the control is biased towards static and linear regulation, ignoring the thermal coupling relationship between heating systems in actual application, which is difficult to achieve high-precision temperature distribution control in complex cavity structure, and is prone to problems such as mutual interference of control signals, system over-regulation or under-regulation, thereby affecting the stability of temperature control and process repeatability. SUMMARY

[0006] In view of the shortcomings of the prior art, the present application provides a PID temperature control system and method for a semiconductor etching machine. To achieve the above object, the present application is implemented by the following technical scheme: a PID temperature control system for a semiconductor etching machine, comprising: A coupling relationship construction module is used to demarcate each heating unit in the static period, each heating unit corresponds to each temperature control area inside the etching cavity of the semiconductor etching machine, and a coupling relationship matrix between each heating unit is constructed. The elements in the coupling relationship matrix are used to represent the influence intensity of temperature change of any temperature control area on temperature change of other temperature control areas.

[0007] A temperature sensing confidence module is used to collect real-time temperature values of each sub-temperature control sensor of each temperature control area in the control period, and after analysis and processing, the temperature real-time sampling values of each temperature control area are obtained and output to the decoupling compensation module.

[0008] A decoupling compensation module is used to generate initial temperature control output values of each temperature control area from a PID controller according to the temperature difference between the preset temperature values of each temperature control area and the corresponding temperature real-time sampling values, and to calculate the actual temperature control output values of each temperature control area based on the coupling relationship matrix and its inverse matrix.

[0009] A coupling relationship dynamic updating module is used to make a micro adjustment to the coupling relationship matrix based on the response residual error between the actual temperature change value and the predicted temperature change value of each temperature control area after the control period ends.

[0010] A transient temperature change control module is used to sense the occurrence of a sharp temperature change event in each temperature control area and perform rapid response processing.

[0011] As a preferred technical scheme, the coupling relationship matrix between each heating unit is constructed, and the specific process is as follows: In the heating unit static state, a unit power step signal is applied to each temperature control area in the etching cavity in turn, and the heating units of other temperature control areas remain unchanged during the application process. After the step signal excitation is applied, the temperature response change curves of each temperature control area are collected respectively, the temperature response coefficients of each temperature control area after the step excitation are summarized, and a coupling relationship matrix is constructed, wherein the elements in the coupling relationship matrix represent the heat influence intensity of the temperature control area to other temperature control areas.

[0012] As a preferred technical solution, the real-time temperature values of each sub-temperature control sensor of each temperature control area are collected, and the temperature real-time sampling values of each temperature control area are obtained after analysis and processing and are output. The specific process is as follows: Each temperature control area is provided with each sub-temperature sensor, and the temperature observation values of each sub-temperature sensor of each temperature control area are collected in a control period. A temperature observation value set is constructed, the temperature observation values in the temperature observation value set are sorted in ascending order, and the temperature observation value at the middle position is selected as the preliminary anti-interference estimated temperature of each temperature control area.

[0013] The other temperature data in the temperature observation value set of the temperature control area are subtracted from the preliminary anti-interference estimated temperature to obtain each preliminary anti-interference temperature difference value of the temperature control area. Based on each preliminary anti-interference temperature difference value, the corresponding temperature fusion weight value is given to each sub-temperature sensor. The temperature observation values of each sub-temperature sensor of the temperature control area are weighted and coupled to obtain the temperature real-time sampling value of the temperature control area. The temperature real-time sampling values of all temperature control areas are counted and output.

[0014] As a preferred technical solution, the initial temperature control output value of each temperature control area is generated by the PID controller, which specifically includes: The preset temperature value of each temperature control area is obtained from the database. The temperature real-time sampling value of each temperature control area is subtracted to obtain the temperature difference value of each temperature control area. Based on the temperature difference value of each temperature control area, the P term, the I term and the D term are obtained by combining the PID three-term control law. The sum of the three terms constitutes the initial temperature control output value of each temperature control area. The initial temperature control output value is used as the output of the outer loop controller and as the set value of the inner loop heating unit controller. The inner loop heating unit controller is used to generate a control voltage signal for driving the heating unit.

[0015] As a preferred technical solution, the actual temperature control output value of each temperature control area is calculated based on the coupling relationship matrix and its inverse matrix, which specifically includes: The initial temperature control output value of each temperature control area generated by the PID controller constitutes a multi-dimensional control instruction set, which is constructed as an initial control vector. The initial control vector includes the initial temperature control output value of each temperature control area in a control period.

[0016] The initial control vector is input into the decoupling compensation module, and the decoupling compensation module is provided with an inverse matrix based on the coupling relationship matrix, and the inverse matrix is used to perform reverse compensation on the cross interference term in the control instruction.

[0017] The initial control vector is multiplied by the inverse matrix to obtain the decoupled actual temperature control output value, and the actual temperature control output value is constructed into an actual control vector.

[0018] As a preferred technical solution, the coupling relationship matrix is adjusted based on the response residual error between the actual temperature change value and the predicted temperature change value of each temperature control area, specifically including: After the control cycle ends, the actual temperature change value of each temperature control area is recorded, and the actual temperature control output value of each temperature control area is combined with the coupling relationship matrix to obtain the predicted temperature change value of each temperature control area.

[0019] The actual temperature change value and the predicted temperature change value of each temperature control area are subtracted to obtain the response residual error of each temperature control area.

[0020] The response residual error vector is obtained based on the response residual error of each temperature control area, and the response residual error vector is used to quantify the prediction error of the coupling relationship matrix in the control cycle. The response residual error vector is associated with the actual control vector, and the corresponding elements in the coupling relationship matrix are modified using the least mean square error.

[0021] As a preferred technical solution, the occurrence of a sharp temperature change event in each temperature control area is perceived and a fast response process is performed, specifically including: In the control cycle, the time sequence of the real-time sampling value of the temperature of each temperature control area is recorded to form a continuous temperature change curve, and the real-time temperature change rate of each temperature control area is obtained based on the real-time temperature slope of the continuous temperature change curve.

[0022] When the real-time temperature change rate of a certain temperature control area exceeds the preset real-time temperature change rate threshold, the temperature control area is recorded as a transient temperature change area, and a fast response process is performed on the transient temperature change area.

[0023] Meanwhile, based on the other temperature control areas whose corresponding coupling row in the coupling relationship matrix has a heat influence intensity exceeding the preset heat influence intensity threshold, the transient temperature change influence area of the transient temperature change area is recorded, and a fast response process is performed on the transient temperature change influence area.

[0024] As a preferred technical solution, the fast response process is performed on the transient temperature change area, specifically including: Switching the control task of the transient temperature change region from the conventional control channel to the high-priority response channel, inputting the temperature change rate difference between the real-time temperature change rate of the transient temperature change region and the real-time temperature change rate threshold into the mapping set of the preset temperature change rate difference-refresh frequency adjustment value in the database, mapping matching to obtain the refresh frequency adjustment value of the high-priority response channel, and applying to the high-priority response channel.

[0025] At the same time, input the temperature change rate difference into the mapping set of the preset temperature change rate difference-PID controller adjustment value set in the database for mapping matching to obtain the PID controller adjustment value set of the heating unit corresponding to the transient temperature change region, and the PID controller adjustment value set includes the proportional term adjustment coefficient, the integral term adjustment coefficient and the differential term adjustment coefficient. Input the heating unit corresponding to the transient temperature change region, and adjust the proportional term, the integral term and the differential term of the PID controller of the transient temperature change region to enhance the error response speed.

[0026] As a preferred technical solution, the transient temperature change influence region is processed quickly, specifically including: Extract the thermal influence intensity value of each transient temperature change influence region and the corresponding transient temperature change region from the coupling relationship matrix, and perform difference processing with the thermal influence intensity threshold to obtain the thermal influence intensity difference value of each transient temperature change influence region. Input the thermal influence intensity difference value of each transient temperature change influence region into the mapping set of the pre-stored thermal influence intensity difference-real-time temperature change rate threshold adjustment factor in the database for mapping matching to obtain the real-time temperature change rate threshold adjustment factor of each transient temperature change influence region, and combine with the preset real-time temperature change rate threshold to obtain the adjusted real-time temperature change rate threshold of each transient temperature change influence region.

[0027] Extract the heating unit physical characteristics of each transient temperature change influence region, including the maximum power capacity, the heating unit thermal inertia and the control cycle length, extract the heating unit physical characteristic scalar set from the database, including the maximum power capacity scalar, the heating unit thermal inertia scalar and the control cycle length scalar, and perform weighted coupling processing after comparing the heating unit physical characteristics of each transient temperature change influence region with the heating unit physical characteristic scalar set to obtain the power change limitation factor of each transient temperature change influence region. Limit the power change of the heating unit of each transient temperature change influence region based on the power change limitation factor of each transient temperature change influence region. The power change limitation factor is used to set the maximum allowable change rate of the power of the heating unit.

[0028] The PID temperature control method for a semiconductor etching machine comprises: In the static period, each heating unit is demarcated, the heating units correspond to each temperature control area inside the etching cavity of the semiconductor etching machine respectively, a coupling relationship matrix between the heating units is constructed, and elements in the coupling relationship matrix are used to represent the influence intensity of temperature change of any temperature control area on temperature change of other temperature control areas.

[0029] In the control period, real-time temperature values of each sub-temperature control sensor of each temperature control area are collected, and temperature real-time sampling values of each temperature control area are obtained after analysis and processing and are output to the decoupling compensation module.

[0030] In the control period, initial temperature control output values of each temperature control area are generated by the PID controller according to temperature differences between preset temperature values of each temperature control area and corresponding temperature real-time sampling values, and actual temperature control output values of each temperature control area are calculated based on the coupling relationship matrix and the inverse matrix thereof.

[0031] After the control period ends, the coupling relationship matrix is adjusted slightly based on response residuals between actual temperature change values and predicted temperature change values of each temperature control area.

[0032] In the control period, a rapid response process is performed in response to a sharp temperature change event occurring in each temperature control area.

[0033] Compared with the prior art, embodiments of the present application have at least the following beneficial effects: (1) The present application provides a PID temperature control system for a semiconductor etching machine, which constructs a coupling relationship matrix in a static period of a heating unit, accurately quantifies thermal coupling relationships between different temperature control areas in an etching cavity, and provides a basis for real-time decoupling compensation in a subsequent temperature control process. The inverse matrix is used for compensation operation, which can effectively eliminate control cross interference caused by heat conduction and heat diffusion between regions, so that heating adjustment of a certain temperature control area will not cause miscontrol of other regions, thereby realizing true "independent controllability". In a semiconductor etching process with micron-level or even nanometer-level precision, subtle temperature differences will cause errors that cannot be ignored. Therefore, high-precision partition temperature control is crucial and can significantly improve etching quality and consistency.

[0034] (2) The present application uses a multi-sub-temperature sensor fusion mechanism, arranges multiple sensing points in each temperature control area, and uses a method based on median value screening and weight weighting to fuse temperature measurement results. This not only effectively eliminates the interference of local abnormal sensor readings on overall temperature control judgment, but also improves the robustness of the system to external fluctuations, process disturbances and other factors. Temperature real-time sampling values obtained by combining the anti-interference estimation strategy become important input data for PID control and subsequent decoupling compensation, ensuring the stability and temperature measurement reliability of the system under complex working conditions.

[0035] (3) The present application forms a response residual by recording the deviation between the actual temperature change and the predicted value after each control cycle, and dynamically fine-tunes the coupling relationship matrix by adopting the least mean square error strategy in combination with the control vector. This adaptive modeling method can continuously optimize the accuracy of the thermal coupling model, offset the modeling errors caused by system aging, environmental changes or process parameter drift, and enable the system to have a certain degree of "self-learning" ability. This dynamic calibration mechanism improves the long-term precision and operation adaptability of the control system, and is especially suitable for the scene of long process cycle and frequent parameter fluctuation in semiconductor manufacturing.

[0036] (4) The present application detects transient temperature change, when the temperature change rate of a certain area exceeds the preset threshold, the system immediately increases the control refresh frequency, and automatically calls the PID adjustment parameters matched with the temperature change rate difference, to quickly respond to the target area temperature. At the same time, the area will also be automatically identified and the threshold will be adjusted accordingly. This mechanism significantly improves the adaptive ability and response speed of the system to sudden thermal interference, effectively guarantees the stability and process safety of the cavity temperature control.

[0037] Of course, any product implementing the present application does not necessarily need to achieve all the above advantages at the same time. BRIEF DESCRIPTION OF DRAWINGS

[0038] Figure 1 The figure is a schematic diagram of the system module of the present application.

[0039] Figure 2 The figure is a schematic diagram of the method flow of the present application.

[0040] Figure 3 The figure is a schematic diagram of the logic flow of the present application.

[0041] Figure 4 The figure is an etching process management system etching process monitoring interface diagram involved in the embodiment of the present application.

[0042] Figure 5 The figure is an etching process management system etching process monitoring interface diagram involved in the embodiment of the present application.

[0043] Figure 6 The figure is a temperature control equipment management interface diagram of the etching process management system involved in the embodiment of the present application. DETAILED DESCRIPTION

[0044] With reference to the accompanying drawings, the technical solutions in the embodiments of the present application will be clearly and completely described below, obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work belong to the protection scope of the present application.

[0045] In the description of the present application, it should be understood that the terms "opening", "upper", "lower", "thickness", "top", "middle", "length", "inner", "periphery" and the like indicate the orientation or positional relationship, only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the components or elements referred to must have a particular orientation, be constructed and operated in a particular orientation, therefore cannot be understood as a limitation of the present application.

[0046] Please refer to Figure 1 As shown in the figure, the PID temperature control system for the semiconductor etching machine provided by the embodiments of the present application specifically comprises: As Figure 3 The logical flowchart involved in the embodiments of the present application is shown, which is used to represent the logical response flow of the present application, so as to intuitively understand the overall working flow and the logical relationship between the steps of the present application. There are two main logical lines: The main control flow starts from the heating unit static state, first constructs the coupling relationship matrix, then collects the sub-temperature sensor data and calculates the real-time sampling value. After calculating the temperature difference of each region, the initial control output is generated, and the thermal interference between regions is eliminated through decoupling compensation processing. The output control voltage signal drives the heating unit to execute control, records the temperature change, predicts the temperature change value of the next period, calculates the response residual and corrects the coupling relationship matrix, forming a continuously adaptive closed loop control.

[0047] After calculating the temperature change rate, the system will judge whether it exceeds the set threshold. If it exceeds the threshold, the region is marked as a transient temperature change region, immediately switches to a high-priority channel, and adjusts the refresh frequency and PID parameters to quickly respond to abnormal fluctuations. At the same time, it will also determine other temperature control regions affected by it, dynamically adjust the temperature change rate threshold of these regions, and avoid secondary interference diffusion.

[0048] The coupling relationship construction module is used to delimit each heating unit in the static period, and each heating unit corresponds to each temperature control region inside the etching cavity of the semiconductor etching machine, to construct the coupling relationship matrix between each heating unit. The elements in the coupling relationship matrix are used to represent the influence intensity of the temperature change of any temperature control region on the temperature change of other temperature control regions.

[0049] With the heating unit in a static state, a unit power step signal (e.g., instantaneously increasing the output power of a heating unit by 20W) is applied sequentially to each temperature-controlled region in the etching cavity according to its region number. During this process, the heating units in other temperature-controlled regions remain unchanged. After the step signal excitation is applied, the temperature response curves of each temperature-controlled region are collected. The temperature responses of each temperature-controlled region after the step excitation are summarized. The purpose of applying the step signal is to simulate the disturbance effect of a single temperature increase in a temperature-controlled region on the thermal field of the entire system, constructing a coupling relationship matrix. This coupling relationship matrix is ​​used to quantitatively describe the influence intensity of heating one region on the temperature of other regions, with units of degrees Celsius per watt (°C / W). The specific calculation process is as follows: Where h{i,j} represents the intensity of the thermal influence of the i-th temperature control zone on the j-th temperature control zone. Let j be the temperature change value of the j-th temperature control zone. This represents the unit power increase value of the i-th temperature control zone. The coupling matrix is ​​a two-dimensional numerical representation matrix that fully reflects the thermodynamic coupling behavior between the heating units of the system.

[0050] The temperature sensing confidence module is used to collect the real-time temperature values ​​of each sub-temperature control sensor in each temperature control zone during the control cycle, analyze and process them to obtain the real-time temperature sampling value of each temperature control zone, and output it to the decoupling compensation module.

[0051] Each temperature control zone is equipped with a sub-temperature sensor. During the control cycle, the temperature observation values ​​of each sub-temperature sensor in each temperature control zone are collected to construct a temperature observation value set. The temperature observation values ​​in the temperature observation value set are sorted in ascending order, and the temperature observation value at the middle position is selected as the preliminary anti-interference estimated temperature for each temperature control zone. It should be noted that if there are two temperature observation values ​​at the middle position in the temperature observation value set, they are averaged and used as the preliminary anti-interference estimated temperature.

[0052] The initial anti-interference temperature difference values ​​for each temperature control area are obtained by subtracting other temperature data from the temperature observation set of the temperature control area and the initial anti-interference estimated temperature. Based on each initial anti-interference temperature difference value, a corresponding temperature fusion weight value is assigned to each sub-temperature sensor. The specific process includes: Each initial anti-interference temperature difference value is input into a pre-stored mapping set of initial anti-interference temperature difference values ​​and temperature fusion weight values ​​in the database for mapping and matching to obtain the temperature fusion weight value of each sub-temperature sensor. The mapping set of initial anti-interference temperature difference values ​​and temperature fusion weight values ​​in the database can be trained or optimized according to the actual equipment operation conditions, thereby adapting to the performance characteristics and application environment of different batches of sensors.

[0053] The temperature real-time sampling values of the temperature control areas are obtained by weighting and adding the temperature observation values of the temperature sensors in the temperature control areas, and the temperature real-time sampling values of all the temperature control areas are counted and output.

[0054] It should be noted that in the multi-sensor temperature measurement scene, individual temperature sensors may have sudden values or noise values due to electromagnetic interference, local environmental changes or their own faults. The median value is the middle value after sorting the observation value set, so even if there are extreme deviated data, it will not significantly affect the calculation result of the median value, which makes it more optimal in anti-interference than the average value. In the actual industrial environment, the data measured by the temperature sensor is often not an ideal normal distribution, but may present a certain skewness or multi-peak structure. As a non-parametric statistical quantity, the median value does not depend on the distribution form of the data and can more stably reflect the central tendency of the data. Compared with complex filtering algorithms or fitting algorithms, the extraction of the median value only needs to perform one sorting and simple judgment, which is very suitable for the scene with short control period and high real-time requirement, especially for the temperature closed-loop control of industrial equipment such as etching machines.

[0055] As Figure 4 The etching process management system in the embodiment of the present application involves an etching process monitoring interface diagram. Specifically, temperature monitoring includes temperature control area display diagram, area number, real-time temperature, set temperature, temperature difference, control output value and state. At the same time, operation control UI is provided to facilitate manual control. Helps operators to master the temperature distribution and control effect in the etching cavity in real time, and at the same time supports manual intervention through the operation control UI, to ensure accurate management and flexible adjustment of the temperature control process.

[0056] The decoupling compensation module is configured to generate initial temperature control output values of the temperature control areas according to the temperature differences between the preset temperature values of the temperature control areas and the corresponding temperature real-time sampling values, and calculate actual temperature control output values of the temperature control areas based on the coupling relationship matrix and its inverse matrix.

[0057] The preset temperature value of each temperature control area is obtained from the database, and the temperature difference value of each temperature control area is obtained after the real-time sampling value of the temperature of each temperature control area is subtracted. Based on the temperature difference value of each temperature control area, combined with the PID three-term control law, the P term (proportional term), I term (integral term) and D term (derivative term) are obtained. The proportional term reflects the influence of the current error itself on the adjustment. After the temperature difference value is input, it is multiplied by the pre-set proportional gain coefficient to obtain the proportional term. The integral term accumulates the history error for eliminating the steady-state deviation. After the history temperature difference value is multiplied by the time interval and accumulated, it is multiplied by the integral gain coefficient to obtain the integral term. The derivative term predicts the error change trend, plays a role in suppressing overshoot and responding in advance. After the current temperature difference value is subtracted from the temperature difference value of the last control period and divided by the control period time length, it is multiplied by the derivative gain coefficient to obtain the derivative term.

[0058] The sum of the three terms constitutes the initial temperature control output value of each temperature control area. The initial temperature control output value is used as the output of the outer loop controller and as the set value of the inner loop heating unit controller. The inner loop heating unit controller is used to generate a control voltage signal for driving the heating unit. It should be noted that in the temperature control system, after the initial temperature control output value of each temperature control area is input to the PID controller, the corresponding control voltage signal can be obtained, because the core function of the PID controller is to convert the temperature error into an executable physical signal for driving the heating unit to perform actual adjustment. The initial control output value is composed of proportional (P), integral (I) and derivative (D) three terms, which comprehensively reflects the current error size, error history accumulation and error change rate. After receiving this control quantity, the PID controller will convert it into a continuously adjustable voltage output signal. Common forms include analog voltage signals, PWM duty cycle signals or adjustable digital control signals, depending on the hardware interface type of the system. The control voltage signal directly acts on the power control device corresponding to the heating unit, such as thyristor, solid-state relay or linear drive module, to adjust the size and duration of its output power, thereby realizing dynamic temperature control of the heating area. Through this process, the system can realize a closed-loop control mechanism from temperature difference judgment to actual energy adjustment, ensuring that the temperature of each area is accurately and stably maintained near the target value.

[0059] The initial temperature control output value of each temperature control area generated by the PID controller constitutes a multi-dimensional control instruction set, which is constructed as an initial control vector, specifically, the set is converted into a column vector, and the initial control vector includes the initial temperature control output value of each temperature control area in the control period. It should be noted that in the embodiment of the present application, the PID controller of each temperature control area independently outputs an initial temperature control output value according to the current temperature difference, which represents the adjustment strength required for the area. The initial temperature control output values of all temperature control areas are collected together to obtain an initial temperature control output value set arranged in a fixed order, wherein each element corresponds to a specific area. Such a set is essentially an ordered numerical sequence with a clear physical meaning, and has the mathematical properties and physical meaning of a vector.

[0060] Figure 6 The temperature control device management interface of the etching process management system is shown in the embodiment of the present application. The PID parameter configuration module can manually configure the coefficients of the three items in the PID by selecting the temperature control area, and provides two control modes, namely manual control and automatic control, which can be selected according to the requirements to realize personalized configuration and optimization of the temperature control strategy and improve the adaptability and control accuracy of the system.

[0061] The initial control vector is input into the decoupling compensation module, and the inverse matrix based on the coupling relationship matrix is provided in the decoupling compensation module. The inverse matrix is used to compensate the cross interference term in the control instruction in the reverse direction. In the control system, the significance of the inverse matrix is to offset the coupling effect represented by the original matrix. Multiplying the control vector by the inverse matrix can compensate the control instruction in the reverse direction, thereby reducing the influence of cross interference.

[0062] The initial control vector and the inverse matrix are subjected to matrix multiplication to obtain the actual temperature control output value after decoupling, and the actual temperature control output value is constructed as an actual control vector. For example, there are two temperature control areas, and the coupling relationship matrix between the two temperature control areas is as follows:

[0063] Wherein, the element in the first row and the first column represents that the first temperature control area heating will cause a heat influence intensity of 1℃ / W to itself, the element in the first row and the second column represents that the first temperature control area heating will cause a heat influence intensity of 0.2℃ / W to the second temperature control area, the element in the second row and the second column represents that the second temperature control area heating will cause a heat influence intensity of 1℃ / W to itself, and the element in the second row and the first column represents that the second temperature control area heating will cause a heat influence intensity of 0.3℃ / W to the first temperature control area.

[0064] Suppose the two temperature control areas calculate that they need to heat 10W and 15W respectively through the PID controller in this control cycle. The initial temperature control output values of the PID controller outputs of the two temperature control areas are 10 and 15 respectively, which constitute the initial control vector, that is:

[0065] To eliminate the coupling interference, the initial control vector is multiplied by the inverse matrix. The specific process is as follows: Calculate the inverse matrix of the coupling relationship matrix A

[0066] It should be noted that in this example, the precision is four decimal places.

[0067] The actual temperature control output values are obtained by multiplying the initial control vector by the inverse matrix:

[0068] Wherein is the actual control vector, is the actual temperature control output value of the first temperature control area, is the actual temperature control output value of the second temperature control area. It should be explained that in the actual system, the actual temperature control output value needs to be converted into a control signal that can be executed in the physical world. The control hardware (such as power amplifier, digital PWM controller) usually has a fixed minimum resolution unit. In this example, the minimum adjustment step of the controller is 0.5W, so 12.766 will be rounded to the nearest executable value, that is, 13W. This quantization processing is a common engineering implementation method in industrial systems.

[0069] In this example, after operation, the decoupled actual control output values are 7.5W and 13W respectively, which are input into the control circuit of the corresponding heating unit and drive heating through voltage regulation.

[0070] The coupling relationship dynamic updating module is used to make a micro adjustment to the coupling relationship matrix based on the response residual error between the actual temperature change value and the predicted temperature change value of each temperature control area after the control cycle ends.

[0071] After the control cycle ends, the actual temperature change value of each temperature control area is recorded, and the predicted temperature change value of each temperature control area is calculated based on the actual temperature control output value of each temperature control area and the coupling relationship matrix.

[0072] The response residual error of each temperature control area is obtained by subtracting the actual temperature change value from the predicted temperature change value.

[0073] ​The response residual vector is obtained based on the response residuals of each temperature control region, and is used to quantify the prediction error of the coupling relationship matrix in the control period. The response residual vector is associated with the actual control vector, and the corresponding elements in the coupling relationship matrix are modified using the least mean square error.

[0074] It should be noted that in order to modify the coupling relationship matrix, the source of the residual needs to be analyzed, so the actual control output values of each temperature control region in the control period are combined into an actual control vector, which is associated with the response residual vector. Specifically, this association is achieved by constructing an error optimization model based on the least mean square (LMS) criterion. The predicted temperature change vector of the coupling relationship matrix can be obtained by multiplying the coupling relationship matrix and the actual control vector. When the response residual occurs, it is determined that there is an error element in the coupling relationship matrix. Based on the least mean square error, the sum of the squares of the residuals is minimized, and the objective function is where is the sum of the squares of the response residuals, is the response residual, and is the difference between the actual temperature change value and the predicted temperature change value of each temperature control region, and the predicted temperature change value is obtained by multiplying the coupling relationship matrix and the actual control vector. Therefore, the position of the error element in the coupling relationship matrix is deduced in reverse.

[0075] Continuing with the example mentioned above, , the actual control output values are 7.5W and 13W, but in this example, for convenience of calculation, still use, it should be noted that in actual application, the error between the theoretical value and the actual control output value needs to be considered. Assuming that the actual temperature change values of the two temperature control regions are 8℃ and 13℃, respectively, denoted as:

[0076] The calculation process of the predicted temperature change value is as follows:

[0077] It should be noted that the self-thermal response strength of the two temperature control regions is known to be 1, i.e. outputting 1W of power exactly raises 1℃, so the adjusted predicted temperature change after decoupling compensation exactly meets the ideal temperature control state of the initial temperature control output value, which proves the effect of decoupling compensation from a mathematical point of view. It should also be noted that for the convenience of subsequent response residual calculation and coupling relationship matrix fine-tuning, the temperature change value is rounded to an integer, and in actual application, the calculation precision needs to be determined according to actual requirements.

[0078]

[0079] Let the perturbation of error elements be z, then:

[0080]

[0081] Optimize the elements in the coupling relationship matrix A, so that is minimum.

[0082] It needs to be explained that in this example, the heat influence intensity of the two temperature control areas is -2, taking the first temperature control area as an example, two error elements can be deduced: The heat influence intensity of the first temperature control area on itself is too large, that is, the 1 element is too large, and the actual temperature of the first temperature control area is less than 1℃ when the control output of 1W is input.

[0083] The heat influence intensity of the second temperature control area on the first temperature control area is too large, that is, the 0.2 element is too large, and the actual temperature of the first temperature control area is less than 0.2℃ when the control output of 1W is input.

[0084] Then there are two corresponding adjustment strategies, which include: ; and ; Wherein, is the system preset adjustment rate factor.

[0085] The transient temperature change control module is used to perceive the dramatic temperature change event occurring in each temperature control area in the control period and perform fast response processing.

[0086] In the control period, the time sequence record of the real-time sampling value of the temperature of each temperature control area is formed to form a continuous temperature change curve, and the real-time temperature change rate of each temperature control area is obtained based on the real-time temperature slope of the continuous temperature change curve.

[0087] When the real-time temperature change rate of a certain temperature control area exceeds the preset real-time temperature change rate threshold, the temperature control area is recorded as a transient temperature change area, and the transient temperature change area is processed with fast response.

[0088] Meanwhile, based on the corresponding coupling row in the coupling relationship matrix, other temperature control areas with heat influence intensity exceeding the preset heat influence intensity threshold are recorded as transient temperature change influence areas of the transient temperature change area, and the transient temperature change influence areas are processed with fast response.

[0089] The fast response processing of the transient temperature change area includes: Switching the control task of the transient temperature change region from the conventional control channel to the high-priority response channel, inputting the temperature change rate difference between the real-time temperature change rate of the transient temperature change region and the real-time temperature change rate threshold into the mapping set of the preset temperature change rate difference-refresh frequency adjustment value in the database, mapping matching to obtain the refresh frequency adjustment value of the high-priority response channel, and applying to the high-priority response channel.

[0090] At the same time, input the temperature change rate difference into the mapping set of the preset temperature change rate difference-PID controller adjustment value set in the database for mapping matching to obtain the PID controller adjustment value set of the heating unit corresponding to the transient temperature change region, and the PID controller adjustment value set includes the proportional term adjustment coefficient, the integral term adjustment coefficient and the differential term adjustment coefficient. Input the heating unit corresponding to the transient temperature change region, and adjust the proportional term, the integral term and the differential term in the PID controller of the transient temperature change region to enhance the error response speed.

[0091] It should be noted that, in the fast response processing process, specifically, the original proportional term parameter is multiplied by the proportional term adjustment coefficient to obtain a new proportional term coefficient for enhancing the instantaneous response ability of the controller to the current temperature error. The integral term parameter is multiplied by the integral term adjustment coefficient to adjust the accumulation rate of the temperature deviation of the system to avoid the problems of excessive integration or response lag. The differential term parameter is multiplied by the differential term adjustment coefficient to improve the prediction ability of the controller to the temperature change trend, so as to take brake control in advance when the temperature is about to change dramatically, and suppress the system overshoot.

[0092] After completing the PID parameter adjustment, the new control parameter will be immediately applied to the heating unit corresponding to the temperature control region. Through this process, the sudden temperature change can be responded more timely and accurately, the chain disturbance caused by temperature control abnormality can be effectively suppressed, and the dynamic response performance and overall stability of the temperature control system can be improved.

[0093] The fast response processing of the transient temperature change influence region specifically includes: Extracting the thermal influence intensity value of each transient temperature change influence region and the corresponding transient temperature change region from the coupling relationship matrix, and performing difference processing on the thermal influence intensity threshold to obtain the thermal influence intensity difference value of each transient temperature change influence region. Input the thermal influence intensity difference value of each transient temperature change influence region into the mapping set of the pre-stored thermal influence intensity difference-real-time temperature change rate threshold adjustment factor in the database for mapping matching to obtain the real-time temperature change rate threshold adjustment factor of each transient temperature change influence region, and multiply the preset real-time temperature change rate threshold to obtain the adjusted real-time temperature change rate threshold of each transient temperature change influence region. The greater the difference, the lower the real-time temperature change rate threshold.

[0094] The physical characteristics of the heating unit in each transient temperature change influence area are extracted, including the maximum power capacity, the heating unit thermal inertia, and the control cycle length. The maximum power capacity refers to the maximum electric power that the heating unit can withstand or output, usually measured in watts (W). It limits the maximum power output of the heating unit, preventing equipment overload or damage, and determines the heating capacity of the heating unit and the limit speed of temperature rise and fall. The heating unit thermal inertia represents the degree of response delay of the unit to temperature changes, reflected as the delay in temperature rise or fall, measured in seconds. A heating unit with greater thermal inertia changes temperature more slowly, so a more gradual power adjustment is needed when controlling to avoid temperature fluctuations or overshoot, ensuring temperature stability. The control cycle length refers to the time interval for the temperature control system to collect sensor data, calculate control instructions, and output control signals, usually measured in milliseconds or seconds. The shorter the control cycle, the faster the system responds, allowing for more timely adjustment of heating power; the longer the cycle, the slower the response, but the system runs more stably and has less computational pressure.

[0095] A set of heating unit physical characteristic scalars is extracted from the database, including the maximum power capacity scalar, the heating unit thermal inertia scalar, and the control cycle length scalar. After comparing and weighting the physical characteristics of the heating unit in each transient temperature change influence area with the set of heating unit physical characteristic scalars, the power change limiting factor of each transient temperature change influence area is obtained through coupling processing. The specific process includes: ; wherein, is the power change limiting factor of the pth transient temperature change influence area, is the maximum power capacity of the pth transient temperature change influence area, is the heating unit thermal inertia of the pth transient temperature change influence area, is the control cycle length of the pth transient temperature change influence area, is the maximum power capacity scalar, is the heating unit thermal inertia scalar, is the control cycle length scalar, is the maximum power capacity weighting factor, is the heating unit thermal inertia weighting factor, is the control cycle length weighting factor, and p is the transient temperature change influence area number, p = 1, 2, 3,..., n, n is the total number of transient temperature change influence areas.

[0096] It should be noted that the maximum power capacity weighting factor, the heating unit thermal inertia weighting factor and the control cycle length weighting factor are used to adjust the influence weight of each parameter in the above formula. The maximum power capacity weighting factor reflects the constraint of the heating unit output power limit on the power change limiting factor, the greater the weight, the more cautious the power upper limit will be considered in adjustment to avoid overload. The heating unit thermal inertia weighting factor is used to reflect the slow response characteristics of temperature, the higher the weight, the more gentle the prediction and adjustment of temperature change will be to reduce temperature fluctuation and overshoot. The control cycle length weighting factor affects the sensitivity of temperature sampling and control command update frequency, when the weight is larger, the response adjustment ability to the change of control cycle will be enhanced. The acquisition of these weighting factors is based on actual device parameter testing and historical operation data in the embodiments of the present application, which is analyzed and processed by machine learning algorithm.

[0097] It should also be noted that the three parameters of maximum power capacity, heating unit thermal inertia and control cycle length are related to each other in the temperature control system. The maximum power capacity determines the upper limit of the output power of the heating unit, which limits the amount of heat that can be provided by the heating unit in unit time. The thermal inertia of the heating unit reflects the hysteresis characteristics of the temperature response of the heating unit, that is, the speed and stability of temperature change after heat input, and the heating unit with large thermal inertia responds slowly to power change. The control cycle length affects the sampling and adjustment frequency of the temperature control system, and a shorter control cycle can respond faster to temperature change, but it may also lead to frequent changes of control command. The relationship between the three is that, under the condition that the maximum power capacity is fixed, the thermal inertia of the heating unit determines the actual effect of temperature change after power adjustment. Therefore, the three parameters need to be considered comprehensively.

[0098] The power change limiting factor based on the influence area of each transient temperature change limits the power change of the heating unit in each transient temperature change influence area, and the power change limiting factor is used to set the maximum allowable change rate of the power of the heating unit.

[0099] Figure 5 For the etching process management system involved in the embodiments of the present application, the etching process monitoring interface diagram is continued. It includes a transient temperature change response module, which displays the transient temperature change area and the affected area, displays the response channel refresh parameters and PID temporary adjustment parameters, and provides a manual setting area for temperature ramp threshold and thermal influence intensity threshold. The transient temperature change area and its affected area recognized by the system are highlighted, the refresh frequency of the response channel and the PID temporary adjustment parameters are displayed in real time, and the manual setting interface of the temperature ramp threshold and the thermal influence intensity threshold are provided, so that the operator can monitor and adjust the rapid response ability of the system to sudden temperature change, and ensure the stability and efficiency of the temperature control system.

[0100] In the embodiment, the application provides a PID temperature control method for a semiconductor etching machine, which specifically comprises the following steps: Figure 2 As shown in the figure, specifically comprising: During the static period, each heating unit is demarcated, the heating units correspond to each temperature control area inside the etching cavity of the semiconductor etching machine respectively, a coupling relationship matrix between the heating units is constructed, and elements in the coupling relationship matrix are used to represent the influence intensity of temperature change of any temperature control area on temperature change of other temperature control areas.

[0101] During the control period, real-time temperature values of each sub-temperature control sensor of each temperature control area are collected, and temperature real-time sampling values of each temperature control area are obtained after analysis and processing and then output to the decoupling compensation module.

[0102] During the control period, initial temperature control output values of each temperature control area are generated by the PID controller according to temperature differences between preset temperature values of each temperature control area and corresponding temperature real-time sampling values, and actual temperature control output values of each temperature control area are calculated based on the coupling relationship matrix and the inverse matrix thereof.

[0103] After the control period ends, the coupling relationship matrix is adjusted slightly based on response residuals between actual temperature change values and predicted temperature change values of each temperature control area.

[0104] During the control period, a rapid response process is performed on the basis of a sharp temperature change event occurring in each temperature control area.

[0105] It should be noted that, in this document, relational terms such as first and second and the like can only be used to distinguish one entity or action from another entity or action, without necessarily requiring or implying any actual such relationship or order between or among the entities or actions. Moreover, the terms "comprises", "comprising", or any other variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can include other elements not expressly listed or inherent to such process, method, article, or apparatus.

[0106] The preferred embodiments of the application disclosed above are only used to help explain the application. The preferred embodiments do not describe all of the details of the application, nor limit the application to the specific embodiments. Obviously, many modifications and variations can be made to the application based on the content of the specification. The embodiments are selected and described in the specification to better explain the principles and practical application of the application, so that those skilled in the art can better understand and utilize the application. Any modifications and variations made to the application without departing from the structure of the application or exceeding the scope defined by the application shall fall within the protection scope of the application.

Claims

1. A PID temperature control system for a semiconductor etching machine, characterized by, The application relates to a semiconductor etching chamber temperature control method and device. The application comprises the following: A coupling relationship construction module is used for demarcating heating units in a static period, the heating units respectively correspond to temperature control areas in an etching chamber of a semiconductor etching machine, a coupling relationship matrix among the heating units is constructed, and elements in the coupling relationship matrix are used for representing the influence intensity of temperature change of any temperature control area on temperature change of other temperature control areas; A temperature sensing confidence module is used for collecting real-time temperature values of each sub-temperature control sensor of each temperature control area in a control period, obtaining temperature real-time sampling values of each temperature control area after analysis and processing, and outputting the temperature real-time sampling values to a decoupling compensation module; The decoupling compensation module is used for generating initial temperature control output values of each temperature control area by a PID controller according to temperature differences between preset temperature values of each temperature control area and corresponding temperature real-time sampling values, and calculating actual temperature control output values of each temperature control area based on the coupling relationship matrix and an inverse matrix thereof; A coupling relationship dynamic updating module is used for performing micro-adjustment on the coupling relationship matrix based on response residuals between actual temperature change values and predicted temperature change values of each temperature control area after the control period ends; 2. The PID temperature control system for a semiconductor etching machine of claim 1, wherein: A transient temperature change control module is used for sensing a sharp temperature change event occurring in each temperature control area in the control period and performing rapid response processing. The coupling relationship matrix among the heating units is constructed in the following specific process:

3. The PID temperature control system for a semiconductor etching machine of claim 1, wherein: In a static state of the heating units, unit power step signals are sequentially applied to each temperature control area in the etching chamber, the heating units of other temperature control areas remain unchanged during the application process, temperature response change curves of each temperature control area are collected after the step signal excitation application, temperature response coefficients of each temperature control area after the step excitation application are summarized, and the coupling relationship matrix is constructed, wherein elements in the coupling relationship matrix represent the thermal influence intensity of the temperature control area on other temperature control areas. The real-time temperature values of each sub-temperature control sensor of each temperature control area are collected, the temperature real-time sampling values of each temperature control area are obtained after analysis and processing, and the temperature real-time sampling values are outputted in the following specific process: Each sub-temperature sensor is arranged in each temperature control area, temperature observation values of each sub-temperature sensor of each temperature control area are collected in the control period, a temperature observation value set is constructed, the temperature observation values in the temperature observation value set are sorted in ascending order, and a temperature observation value at a middle position is selected as a preliminary anti-interference estimated temperature of each temperature control area; 4. The PID temperature control system for a semiconductor etching machine of claim 1, wherein: Other temperature data in the temperature observation value set of the temperature control area are subtracted from the preliminary anti-interference estimated temperature to obtain each preliminary anti-interference temperature difference value of the temperature control area, corresponding temperature fusion weight values are given to each sub-temperature sensor based on the preliminary anti-interference temperature difference values, the temperature observation values of each sub-temperature sensor of the temperature control area are coupled after weighting, the temperature real-time sampling values of the temperature control area are obtained, and the temperature real-time sampling values of all temperature control areas are counted and outputted. The PID controller generates the initial temperature control output values of each temperature control area, and the specific process comprises the following: The preset temperature value of each temperature control area is obtained from the database, and the temperature difference value of each temperature control area is obtained after the real-time sampling value of the temperature of each temperature control area is subtracted from the preset temperature value. Based on the temperature difference value of each temperature control area, the P term, the I term and the D term are obtained by combining the PID three-term control law. The sum of the three terms constitutes the initial temperature control output value of each temperature control area. The initial temperature control output value is used as the output of the outer loop controller and as the set value of the inner loop heating unit controller. The inner loop heating unit controller is used to generate a control voltage signal for driving the heating unit.

5. The PID temperature control system for a semiconductor etching machine of claim 1, wherein: The actual temperature control output value of each temperature control area is calculated based on the coupling relationship matrix and its inverse matrix, specifically including: The initial temperature control output value of each temperature control area generated by the PID controller constitutes a multi-dimensional control instruction set, which is constructed as an initial control vector. The initial control vector includes the initial temperature control output value of each temperature control area within a control period. The initial control vector is input into the decoupling compensation module, which is provided with an inverse matrix based on the coupling relationship matrix. The inverse matrix is used to perform reverse compensation on the cross interference term in the control instruction. The initial control vector and the inverse matrix are subjected to matrix multiplication to obtain the decoupled actual temperature control output value. The actual temperature control output value is constructed as an actual control vector.

6. The PID temperature control system for a semiconductor etching machine of claim 1, wherein: The coupling relationship matrix is adjusted based on the response residual error between the actual temperature change value and the predicted temperature change value of each temperature control area, specifically including: After the control period ends, the actual temperature change value of each temperature control area is recorded, and the predicted temperature change value of each temperature control area is calculated based on the actual temperature control output value of each temperature control area and the coupling relationship matrix. The actual temperature change value and the predicted temperature change value of each temperature control area are subtracted to obtain the response residual error of each temperature control area. The response residual error vector is obtained based on the response residual error of each temperature control area. The response residual error vector is used to quantify the prediction error of the coupling relationship matrix in the control period. The response residual error vector and the actual control vector are associated, and the corresponding elements in the coupling relationship matrix are modified using the least mean square error.

7. The PID temperature control system for a semiconductor etching machine of claim 1, wherein: The occurrence of a rapid temperature change event in each temperature control area is perceived, and a rapid response process is performed, specifically including: In the control period, the real-time sampling value of the temperature of each temperature control area is recorded in time sequence to form a continuous temperature change curve, and the real-time temperature change rate of each temperature control area is obtained based on the real-time temperature slope of the continuous temperature change curve. When the real-time temperature change rate of a certain temperature control area exceeds the preset real-time temperature change rate threshold, the temperature control area is recorded as a transient temperature change area, and a rapid response process is performed on the transient temperature change area. Other temperature control areas whose corresponding coupling row in the coupling relationship matrix has a heat influence intensity exceeding the preset heat influence intensity threshold are recorded as transient temperature change influence areas of the transient temperature change area, and a rapid response process is performed on the transient temperature change influence areas.

8. The PID temperature control system for a semiconductor etching machine of claim 7, wherein: The rapid response process performed on the transient temperature change area specifically includes: The control task of the transient temperature change region is switched from the conventional control channel to the high-priority response channel, a temperature change rate difference between the real-time temperature change rate of the transient temperature change region and a real-time temperature change rate threshold is input into a mapping set of preset temperature change rate difference-refresh frequency adjustment value in a database, a refresh frequency adjustment value of the high-priority response channel is obtained through mapping matching, and the refresh frequency adjustment value is applied to the high-priority response channel; Meanwhile, the temperature change rate difference is input into a mapping set of preset temperature change rate difference-PID controller adjustment value set in the database for mapping matching to obtain a PID controller adjustment value set of the heating unit corresponding to the transient temperature change region, the PID controller adjustment value set includes a proportional term adjustment coefficient, an integral term adjustment coefficient and a differential term adjustment coefficient, and the proportional term, the integral term and the differential term in the PID controller of the transient temperature change region are adjusted in the heating unit corresponding to the transient temperature change region to enhance the error response speed.

9. The PID temperature control system for a semiconductor etching machine of claim 7, wherein: The fast response processing of the transient temperature change influence region specifically includes: The thermal influence intensity value of each transient temperature change influence region and the corresponding transient temperature change region is extracted from the coupling relationship matrix, and a thermal influence intensity difference value of each transient temperature change influence region is obtained by subtracting the thermal influence intensity threshold value; the thermal influence intensity difference value of each transient temperature change influence region is input into a mapping set of pre-stored thermal influence intensity difference-real-time temperature change rate threshold adjustment factor in the database for mapping matching to obtain a real-time temperature change rate threshold adjustment factor of each transient temperature change influence region, and the real-time temperature change rate threshold adjustment factor is combined with the preset real-time temperature change rate threshold value to obtain an adjusted real-time temperature change rate threshold value of each transient temperature change influence region; The heating unit physical characteristics of each transient temperature change influence region are extracted, including the maximum power capacity, the heating unit thermal inertia and the control cycle length; a heating unit physical characteristic scalar set, including the maximum power capacity scalar, the heating unit thermal inertia scalar and the control cycle length scalar, is extracted from the database; and after comparison and weighted coupling processing of the heating unit physical characteristics of each transient temperature change influence region and the heating unit physical characteristic scalar set, a power change limitation factor of each transient temperature change influence region is obtained, and the power change limitation factor is used to set the maximum allowable change rate of the power of the heating unit.

10. The method applied to the PID temperature control system of the semiconductor etching machine according to any one of claims 1-9, characterized in that: In the static period, each heating unit is demarcated, the heating unit corresponds to each temperature control region inside the etching cavity of the semiconductor etching machine, and a coupling relationship matrix between the heating units is constructed, and the elements in the coupling relationship matrix are used to represent the influence intensity of the temperature change of any temperature control region on the temperature change of other temperature control regions; In the control period, the real-time temperature values of each sub-temperature control sensor of each temperature control region are collected, and after analysis and processing, the temperature real-time sampling values of each temperature control region are output to the decoupling compensation module. In the control period, according to the temperature difference between the preset temperature value and the real-time sampling value of each temperature control area, the initial temperature control output value of each temperature control area is generated by the PID controller, and the actual temperature control output value of each temperature control area is calculated based on the coupling relationship matrix and its inverse matrix; After the end of the control period, the coupling relationship matrix is adjusted slightly based on the response residual between the actual temperature change value and the predicted temperature change value of each temperature control area; In the control period, the rapid response processing is performed by sensing the sharp temperature change event in each temperature control area.

Images