Temperature compensation methods, systems, testing equipment, computer equipment, and storage media

Abstract

This application provides a temperature compensation method, system, detection equipment, computer equipment, and storage medium. The application acquires measurement data of a target component after hardware compensation; processes the measurement data using a pre-calibrated Monte Carlo incremental model to calculate the residual drift value of each sample trajectory at the current step size; reconstructs the true output data based on each residual drift value to obtain a candidate output set; fuses the candidate output set using a preset decision strategy to obtain prediction information; and generates a temperature drift compensation amount based on this prediction information. This application first uses hardware compensation to eliminate most of the linear temperature drift. Based on this, it further employs a Monte Carlo incremental model to accurately predict the temperature drift trajectory at each step size, and then performs real-time compensation on the target component according to the predicted temperature drift trajectory. This eliminates the remaining small portion of nonlinear temperature drift, improves the temperature compensation effect of the target component, and achieves high-precision measurement under wide temperature range variations.

Description

Technical Field

[0001] This application relates to the field of semiconductor testing technology, and in particular to a temperature compensation method, system, testing equipment, computer equipment, and storage medium. Background Technology

[0002] In semiconductor thin film epitaxial growth processes, the wafer surface temperature field is crucial to nucleation kinetics and epitaxial rate, directly affecting growth rate, film quality, doping efficiency, composition ratio, and interface steepness. Epitaxial growth processes operate over a wide temperature range, approximately 500°C to 1200°C, and temperature uniformity has a significant impact on device performance. Multi-zone heating plates require real-time in-situ detection devices for accurate and stable monitoring. Traditionally, a combined detection system (Reflectance-Temperature-Curvature, RTC) is used to achieve this in-situ online detection function. However, components in these combined detection systems commonly exhibit electrical parameter drift with temperature changes, which directly affects measurement accuracy and system stability.

[0003] Negative temperature coefficient (NTC) thermistors are widely used in temperature sensing and compensation systems due to their low cost and high sensitivity. However, the material and structural characteristics of NTCs result in thermal equilibrium hysteresis: when the ambient temperature changes abruptly or rapidly, the heat transfer process cannot be completed instantaneously, and the device temperature response to the ambient temperature exhibits a significant time constant and hysteresis. This, coupled with the device's self-heating effect (caused by bias power dissipation) and the inconsistency between the package / mount thermal resistance, leads to a continuous timing deviation between the measured temperature value and the actual temperature of the target being compensated.

[0004] Traditional compensation techniques mostly focus on static error correction. This correction method relies on fixed parameter mapping relationships and cannot characterize the coupling effect between temperature change rate and response hysteresis. It has a fundamental limitation on the adaptability to accurate detection of changes over a wide temperature range during epitaxial growth processes. Summary of the Invention

[0005] The purpose of this application is to provide a temperature compensation method, system, detection equipment, computer equipment, and storage medium to overcome the limitations of traditional compensation techniques that rely on fixed parameter mapping relationships and cannot accurately detect wide temperature range changes during thin film epitaxial growth processes.

[0006] In a first aspect, this application proposes a temperature compensation method, comprising: acquiring measurement data of a target component at preset step intervals; wherein the measurement data includes temperature data and output voltage compensated by a hardware compensation unit;

[0007] A pre-calibrated Monte Carlo incremental model is used to calculate the residual drift value of each sample trajectory at the current step size based on the measurement data; wherein, each sample trajectory is generated by the Monte Carlo incremental model based on the measurement data;

[0008] Based on the residual drift values ​​of each current step, the true output data is reconstructed to obtain a candidate output set;

[0009] The candidate output set is fused using a preset decision-making strategy to obtain prediction information;

[0010] Based on the predicted information, a temperature drift compensation amount is generated so that the hardware compensation unit can perform temperature compensation on the target component according to the temperature drift compensation amount.

[0011] In one embodiment, the hardware compensation unit includes two symmetrically designed operational amplifiers, and the temperature drift characteristics of the two operational amplifiers are matched.

[0012] In one embodiment, the method for calibrating the Monte Carlo incremental model includes:

[0013] Sampling data of the target component is acquired at preset sampling intervals, wherein the sampling data includes temperature data, output voltage, and reference voltage; the output voltage is obtained after hardware compensation of the target component.

[0014] The residual drift voltage is calculated based on the output voltage and the reference voltage.

[0015] Based on the residual drift voltage and the temperature data, temperature drift characteristic parameters are constructed; wherein, the temperature drift characteristic parameters include at least one of the following: temperature change rate, rate of change of temperature change rate, residual integral, and residual absolute value;

[0016] Based on the temperature drift characteristic parameters, the mean function and standard deviation function of the residual drift increment are constructed to obtain the Monte Carlo incremental model.

[0017] In one embodiment, the mean function includes a temperature variation compensation coefficient; the standard deviation function includes an aging coefficient and a temperature disturbance gain.

[0018] Methods for calibrating Monte Carlo incremental models also include:

[0019] The temperature change compensation coefficient is obtained by measuring the linear regression relationship between the residual and the square of the rate of temperature change under rapid temperature cycling conditions.

[0020] An aging test was conducted at a constant temperature change rate, and a linear relationship was established between the running time and the measured standard deviation function to obtain the aging coefficient.

[0021] The standard deviation of the residuals is measured at different rates of temperature change, and the temperature disturbance gain is determined based on the slope of the residuals as a function of the rate of temperature change.

[0022] In one embodiment, the step of employing a pre-calibrated Monte Carlo incremental model to calculate the residual drift value of each sample trajectory at the current step size based on the measurement data includes:

[0023] Using the Monte Carlo incremental model, the mean and standard deviation of the measurement data for the current step size are calculated; random perturbation is introduced based on the mean and standard deviation to obtain the residual drift increment for the current step size;

[0024] The residual drift increment of each sample trajectory at the current step length is superimposed on the residual drift value of the previous step length to obtain the residual drift value of each sample trajectory at the current step length.

[0025] In one embodiment, the number of sample trajectories is dynamically set, and the number is positively correlated with the absolute value of the temperature change rate in a hyperbolic tangent relationship.

[0026] In one embodiment, the process of reconstructing the true output data based on the residual drift values ​​of each current step to obtain a candidate output set includes:

[0027] For each sample trajectory, the output voltage of the target component is subtracted from the residual drift value of the sample trajectory at the current step size to obtain the true output data;

[0028] The actual output data constitute the candidate output set.

[0029] In one embodiment, the step of fusing the candidate output set using a preset decision strategy to obtain prediction information includes:

[0030] Based on statistical distance metrics, the consistency between each candidate value and its corresponding sample trajectory in the candidate output set and the set distribution is evaluated. Abnormal trajectories and candidate values ​​whose corresponding candidate outputs deviate significantly from the population distribution are identified and excluded to obtain an effective candidate set.

[0031] Based on the similarity between the current temperature change rate and the temperature change rate of each sample trajectory, weights are assigned to the effective candidate set to obtain a weighted candidate set.

[0032] Quantile aggregation is performed on the weighted candidate set to generate a point estimate and confidence interval for the current step size, thus obtaining the prediction information.

[0033] In one embodiment, generating a temperature drift compensation amount based on the predicted information includes:

[0034] The initial compensation amount for the current step size is generated based on the point estimate;

[0035] Based on the confidence interval, the initial compensation amount is adaptively processed to obtain the temperature drift compensation amount.

[0036] In one embodiment, the method further includes:

[0037] Based on the temperature data, nonlinear characteristic parameters are constructed; wherein, the nonlinear characteristic parameters include at least one of the following: temperature change rate, change in temperature change rate, and time decay term of temperature change intensity.

[0038] Based on adaptive weights, the nonlinear characteristic parameters are weighted and summed to obtain the fine-tuning compensation amount.

[0039] Based on the fine-tuning compensation amount, the initial compensation amount after adaptive processing is feedforward corrected to obtain the temperature drift compensation amount.

[0040] In one embodiment, the method further includes:

[0041] A calibration instruction is generated based on the predicted information to adjust the parameters of the hardware compensation unit;

[0042] And / or adjust the parameters of the Monte Carlo incremental model based on the temperature compensation effect of the target element.

[0043] In one embodiment, the method further includes:

[0044] Calculate the rate of temperature change based on the temperature data;

[0045] The temperature change rate is compared with a preset threshold, and the working mode is switched according to the comparison result; wherein, the working mode includes energy-saving mode, high-energy mode and standard mode.

[0046] Secondly, this application proposes a temperature compensation system, the system comprising:

[0047] The acquisition module is used to acquire measurement data of the target component at preset step intervals; wherein, the measurement data includes temperature data and output voltage after hardware compensation;

[0048] The prediction module employs a pre-calibrated Monte Carlo incremental model to calculate the residual drift value of each sample trajectory at the current step length based on the measurement data. Each sample trajectory is generated by the Monte Carlo incremental model based on the measurement data. The module reconstructs the true output data based on the residual drift values ​​at each current step length to obtain a candidate output set. A preset decision strategy is then used to fuse the candidate output set to obtain prediction information.

[0049] The compensation module is used to generate a temperature drift compensation amount based on the predicted information, so that the hardware compensation unit can perform temperature compensation on the target component according to the temperature drift compensation amount.

[0050] Thirdly, this application also provides a testing device, comprising:

[0051] The data acquisition module is used to acquire temperature information of the wafer during the thin film epitaxial growth process.

[0052] A temperature detection module is used to monitor the temperature change of the wafer during the thin film growth process based on the temperature information; the temperature detection module includes a target element and a hardware compensation unit.

[0053] The temperature compensation system described in the second aspect is used to acquire measurement data of the target component and generate a temperature drift compensation amount based on the measurement data, so that the hardware compensation unit can perform temperature compensation on the target component according to the temperature drift compensation amount.

[0054] Fourthly, this application also provides a computer device. The computer device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the method steps of the first aspect.

[0055] Fifthly, this application also provides a computer-readable storage medium. The computer-readable storage medium stores a computer program thereon, which, when executed by a processor, implements the method steps of the first aspect.

[0056] The above-mentioned temperature compensation method, system, testing equipment, computer equipment, and storage medium have at least the following advantages:

[0057] This application acquires measurement data of the target component after hardware compensation; processes the measurement data using a pre-calibrated Monte Carlo incremental model to calculate the residual drift value of each sample trajectory at the current step size; reconstructs the true output data based on each residual drift value to obtain a candidate output set; fuses the candidate output set using a preset decision strategy to obtain prediction information; and generates a temperature drift compensation amount based on this prediction information to compensate the temperature of the target component. This application first uses hardware compensation to eliminate most of the linear temperature drift. Based on this, it further employs a Monte Carlo incremental model to accurately predict the temperature drift trajectory at each step size, and then compensates the target component in real time according to the predicted temperature drift trajectory. This eliminates the remaining small portion of nonlinear temperature drift, improves the temperature compensation effect of the target component, and achieves high-precision measurement under wide temperature range variations. Attached Figure Description

[0058] Figure 1 This is a structural block diagram of the detection device in one embodiment;

[0059] Figure 2 This is a flowchart illustrating a temperature compensation method in one embodiment;

[0060] Figure 3 This is a schematic diagram of the hardware compensation unit in one embodiment;

[0061] Figure 4 This is a structural block diagram of the hardware compensation unit in another embodiment;

[0062] Figure 5 This is a flowchart illustrating the steps for calibrating the Monte Carlo incremental model in one embodiment;

[0063] Figure 6 This is a flowchart illustrating the steps for calculating residual drift values ​​in one embodiment;

[0064] Figure 7 This is a flowchart illustrating the steps of fusing candidate output sets to obtain prediction information in one embodiment;

[0065] Figure 8 This is a structural block diagram of a temperature compensation system in one embodiment;

[0066] Figure 9 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0067] The following specific examples illustrate the implementation of this application. Those skilled in the art can easily understand other advantages and effects of this application from the content disclosed in this specification. This application can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of this application. It should be noted that, unless otherwise specified, the following embodiments and features in the embodiments can be combined with each other.

[0068] Some exemplary embodiments of this application have been described for illustrative purposes. It should be understood that this application may be implemented in other ways not specifically shown in the accompanying drawings.

[0069] Please see Figure 1 , Figure 1 The process chamber 1 contains an organic stage tray 11 and a wafer 12, with the wafer 12 placed on the stage tray 11. Optionally, in one embodiment, this application provides a detection device for detecting the temperature change of the wafer 12 during the thin film growth process in an epitaxial growth process. The detection device includes: a data acquisition module 2, a temperature detection module 3, and a temperature compensation system 4.

[0070] The acquisition module 2, connected to the temperature detection module 3, is used to acquire temperature information of the wafer 12 during the thin film epitaxial growth process and transmit this temperature information to the temperature detection module 3. It should be understood that while the acquisition module 2 in this embodiment is used to acquire temperature information, in practical applications, the acquisition module 2 can also acquire reflectivity and / or curvature as needed. Exemplarily, the acquisition module 2 includes a probe, a transmitting optical fiber, and a receiving optical fiber; wherein the probe is used to coaxially integrate the optical paths of the transmitting and receiving optical fibers and perform fixed-point, fixed-angle coupling on the wafer surface, defining the measurement spot size and working distance; the transmitting optical fiber is used to introduce the detection light source into the cavity and irradiate the wafer surface. The receiving optical fiber is used to collect thermal radiation and reflection signals from the wafer surface and transmit them to the temperature detection module 3.

[0071] Temperature detection module 3, connected to temperature compensation system 4, is used to monitor temperature changes of wafer 12 during thin film growth based on temperature information. Temperature detection module 3 includes target element 31 and hardware compensation unit 32. Target element 31 is a component in the hardware detection chain of temperature detection module 3, and this component may experience electrical parameter drift due to temperature changes.

[0072] The temperature compensation system 4 is used to acquire measurement data of the target component 31 and generate a temperature drift compensation amount based on the measurement data, so that the hardware compensation unit 32 can perform temperature compensation on the target component 31 according to the temperature drift compensation amount. The measurement data includes temperature data and the output voltage after compensation by the hardware compensation unit 32.

[0073] Specifically, the temperature compensation system 4 outputs a temperature drift compensation amount based on the measurement data, including:

[0074] The temperature compensation system acquires measurement data of the target component at preset step intervals. The measurement data includes temperature data and output voltage after compensation by a hardware compensation unit. A pre-calibrated Monte Carlo incremental model is used to calculate the residual drift value of each sample trajectory at the current step based on the measurement data. Each sample trajectory is generated by the Monte Carlo incremental model based on the measurement data. The real output data is reconstructed based on the residual drift value at each current step to obtain a candidate output set. A preset decision strategy is used to fuse the candidate output set to obtain prediction information. The temperature drift compensation amount is generated based on the prediction information.

[0075] The aforementioned testing equipment includes a temperature detection module that monitors temperature changes in the wafer during thin film growth based on temperature information acquired by the acquisition module; a temperature compensation system that acquires measurement data of the target element at preset step intervals, processes the measurement data using a pre-calibrated Monte Carlo incremental model, and calculates the residual drift value of each sample trajectory at the current step; reconstructing the true output data based on each residual drift value to obtain a candidate output set; fusing the candidate output set using a preset decision strategy to obtain prediction information; generating a temperature drift compensation amount based on this prediction information; and a hardware compensation unit performing temperature compensation on the target element based on this temperature drift compensation amount. This application first uses hardware compensation to eliminate most of the linear temperature drift. Based on this, it further employs a Monte Carlo incremental model to accurately predict the temperature drift trajectory at each step, and then performs real-time compensation on the target element based on the predicted temperature drift trajectory, eliminating the remaining small portion of nonlinear temperature drift, improving the temperature compensation effect of the target element, and achieving high-precision measurement of the temperature detection module in environments with wide temperature variations.

[0076] In one exemplary embodiment, this application provides a temperature compensation method, which will be applied to the following embodiments. Figure 1 We will use server 104 as an example to illustrate this.

[0077] Please see Figure 2 , Figure 2 This is a flowchart illustrating a temperature compensation method according to this embodiment, which specifically includes the following steps:

[0078] Step 202: Acquire measurement data of the target component at preset intervals; wherein, the measurement data includes temperature data and output voltage after compensation by the hardware compensation unit.

[0079] Specifically, the preset step size is a discrete time interval used for temperature drift recursion and data sampling. In this embodiment, a preset step size refers to the time required to complete one data measurement, model inference, and trajectory update process. For example, this preset step size Δt can be set according to the sampling frequency, and its expression is: ;in, The sampling frequency is determined by the system clock or the sampling rate of the analog-to-digital converter module to ensure that the model's inference is aligned with the sampling. It should be noted that the preset step size Δt should be greater than or equal to the minimum stable value allowed by the measurement link bandwidth to capture rapid temperature changes of the target component while ensuring that computational resources are manageable.

[0080] Step 204: Using a pre-calibrated Monte Carlo incremental model, calculate the residual drift value of each sample trajectory at the current step length based on the measurement data; wherein, each sample trajectory is generated by the Monte Carlo incremental model based on the measurement data; reconstruct the true output data based on the residual drift value of each current step length to obtain a candidate output set; use a preset decision strategy to fuse the candidate output set to obtain prediction information.

[0081] Specifically, the Monte Carlo method is a numerical computation method that uses random sampling and statistics to solve mathematical, physical, and engineering problems. In this embodiment, the Monte Carlo incremental model makes stepwise predictions of the residual temperature drift of the target component during operation. Within each preset step, it calculates the incremental value by introducing random perturbations, thereby obtaining the residual drift value for the current step.

[0082] The sample trajectory refers to the result obtained by processing the measurement data corresponding to each step using a Monte Carlo incremental model, and then independently sampling the random disturbances multiple times based on the processed result.

[0083] Step 206: Generate temperature drift compensation amount based on the predicted information so that the hardware compensation unit can perform temperature compensation on the target component according to the temperature drift compensation amount.

[0084] The aforementioned temperature compensation method acquires measurement data of the target component after hardware compensation; processes the measurement data using a pre-calibrated Monte Carlo incremental model to calculate the residual drift value of each sample trajectory at the current step length; reconstructs the true output data based on each residual drift value to obtain a candidate output set; fuses the candidate output set using a preset decision strategy to obtain prediction information; and generates a temperature drift compensation amount based on this prediction information to compensate the temperature of the target component. This application first uses hardware compensation to eliminate most of the linear temperature drift. Based on this, it further employs a Monte Carlo incremental model to accurately predict the temperature drift trajectory at each step length, and then compensates the target component in real time according to the predicted temperature drift trajectory. This eliminates the remaining small portion of nonlinear temperature drift, improves the temperature compensation effect of the target component, and achieves high-precision measurement under wide temperature range variations.

[0085] Please see Figure 3 , Figure 3 The diagram shows the structure of the hardware compensation unit. The hardware compensation unit includes two symmetrically designed operational amplifiers, and the temperature drift characteristics of the two operational amplifiers are matched.

[0086] Specifically, the hardware compensation unit includes two operational amplifiers (hereinafter referred to as main operational amplifier U1 and compensation operational amplifier U2). The temperature drift characteristics of main operational amplifier U1 and compensation operational amplifier U2 are matched, and main operational amplifier U1 and compensation operational amplifier U2 are configured as follows: Figure 3The circuit shown has a differential input, single-ended output structure, where the useful signal is the difference between the input and output of the main operational amplifier U1 and the compensation operational amplifier U2. In other embodiments, the main operational amplifier U1 and the compensation operational amplifier U2 can also be configured with a differential input, differential output structure as needed.

[0087] Using the above scheme, the ambient temperature change of the target component ( The drift over time (t) is slow, and the drift voltage / current generated by the main op-amp U1 and the compensation op-amp U2 should be highly correlated and have extremely similar amplitudes and directions under ideal matching. This results in the two highly correlated drift components being largely canceled out as common-mode signals at the output of the hardware compensation unit. The residual temperature drift at the final output will be much smaller than the drift of a single op-amp.

[0088] Optionally, the expression for the temperature drift voltage transfer function of the target component is:

[0089]

[0090] in, Indicates the temperature drift coefficient; Indicates the time drift coefficient; This indicates the temperature-induced acceleration compensation term; This represents the temperature change. This transfer function describes the residual, non-completely eliminated temperature drift voltage after offsetting. It consists of three parts, each corresponding to a different physical mechanism of drift.

[0091] Linear temperature drift term: Even if the main operational amplifier U1 and the compensation operational amplifier U2 are rigorously selected and matched, the temperature-dependent characteristics of their core components cannot be perfectly identical. This slight mismatch leads to the aforementioned linear temperature drift term, which is the primary source of residual temperature drift.

[0092] Time drift item: Semiconductor devices and their internal materials undergo an extremely slow aging process over time, causing drift in their bias voltage and current. Even if two op-amps are perfectly matched at the initial moment, their aging rates may differ slightly.

[0093] Temperature-induced acceleration compensation term: When the ambient temperature changes rapidly, transient temperature differences may exist at different locations inside the operational amplifier chip. Even if the packages are closely adjacent, heat conduction still takes time. This transient spatial temperature gradient will cause a slight difference between the actual temperatures T1(t) and T2(t) experienced by the main operational amplifier U1 and the compensation operational amplifier U2 at the same moment. This transient mismatch drift caused by the thermal gradient is usually proportional to the square of the rate of temperature change dT / dt.

[0094] Generally speaking, ambient temperature compensation can be divided into the following two scenarios.

[0095] In the case of temperature rise, a positive temperature drift voltage +ΔVt is added to the output of the main operational amplifier U1 based on its useful signal. Since the temperature drift characteristic parameters of the main operational amplifier U1 and the compensation operational amplifier U2 are highly matched and they are in the same thermal environment, the output of the main operational amplifier U1 synchronously generates a positive compensation voltage +V't. This compensation voltage +V't is injected into the inverting input of the compensation operational amplifier U2, injecting a positive voltage (+V't) into its inverting input. According to the principle of virtual short and virtual open at the inverting input of the operational amplifier and the characteristics of the inverting amplifier, this causes a negative change -G·V't at the output of the compensation operational amplifier U2, where G is the gain of the compensation operational amplifier U2 from this injection point to its output. The temperature drift signal +ΔVt from the main operational amplifier U1 reaches the inverting input of the compensation operational amplifier U2, generating a reverse compensation current -Icomp. The change in the output of the compensation operational amplifier U2 can be expressed as:

[0096] ΔV_out = -G·(ΔVt - V't) ≈ 0.

[0097] In a temperature drop scenario, a negative temperature drift voltage -ΔVt is added to the output of the main operational amplifier U1 based on its useful signal. For the same reason, the compensation operational amplifier U2 simultaneously generates a negative compensation voltage -V't. The compensation voltage -V't is injected into the inverting input of the compensation operational amplifier U2, generating a reverse compensation current +Icomp. The change in the output of the compensation operational amplifier U2 can be expressed as: ΔV_out = G·(ΔVt - V't)≈ 0

[0098] Using the above scheme, a hardware compensation unit is connected to the output of the target component for hardware compensation, which can eliminate most of the linear temperature drift. However, since the two temperature coefficients cannot be perfectly matched, the final output voltage signal includes a residual temperature drift voltage, denoted as the residual drift voltage. The expression for the residual drift voltage is:

[0099]

[0100] in, , These represent the actual temperature drift coefficients of the main operational amplifier U1 and the compensation operational amplifier U2, respectively. It indicates the amount of change in ambient temperature.

[0101] The final output voltage signal expression is:

[0102]

[0103] Where S(t) is the target voltage at the reference temperature at time t.

[0104] The voltage signal This refers to the output voltage of the aforementioned target component after hardware compensation.

[0105] Please see Figure 4 Optionally, the aforementioned hardware compensation unit may also include a temperature sensor and a compensation digital-to-analog converter.

[0106] A temperature sensor, positioned close to the target component, detects the component's temperature data and transmits it to a temperature compensation system. The temperature compensation system then calculates a temperature compensation value based on the measured data from the target component.

[0107] The compensated digital-to-analog converter, connected to the temperature compensation system and the compensated operational amplifier U2 respectively, is used to convert the digital temperature compensation value into a controllable analog current and inject it into the input of the compensated operational amplifier U2 to offset the residual temperature drift in the output voltage of the target component.

[0108] For example, the expression for the compensation current injected into the input terminal of the compensation operational amplifier U2 is:

[0109]

[0110]

[0111] in, Indicates the compensation coefficient; This represents the equivalent conversion factor of the compensation current to the output. This represents the output voltage after the input voltage Vin passes through the main operational amplifier U1 and the compensation operational amplifier U2.

[0112] Please see Figure 5 Optionally, methods for calibrating the Monte Carlo incremental model include:

[0113] Step 502: Acquire sampling data of the target component at preset sampling intervals. The sampling data includes temperature data, output voltage, and reference voltage; the output voltage is obtained after hardware compensation of the target component.

[0114] Step 504: Calculate the residual drift voltage based on the output voltage and the reference voltage.

[0115] Step 506: Based on the residual drift voltage and temperature data, construct temperature drift characteristic parameters; wherein, the temperature drift characteristic parameters include at least one of the following: temperature change rate, rate of change of temperature change rate, residual integral, and residual absolute value.

[0116] Step S508: Based on the temperature drift characteristic parameters, construct the mean function and standard deviation function of the residual drift increment to obtain the Monte Carlo incremental model.

[0117] Specifically, the preset sampling period is set in the same way as the preset step size mentioned above, and its value can be the same as or different from the preset step size.

[0118] In each preset sampling period, the current sampling data of the target element is acquired.

[0119] The temperature gradient is calculated in real time based on temperature data, and its expression is as follows:

[0120]

[0121] Where k represents the current time; This represents the sampling period, and Δt = 1 / fsample.

[0122] Based on the output voltage and the reference voltage, the expression for the residual drift voltage is calculated as follows:

[0123]

[0124] in, This indicates the target voltage corresponding to the current temperature.

[0125] The expression for the temperature drift characteristic parameter is: X = [ , , , . Indicates the rate of temperature change. This represents the rate of change of temperature. Represents the residual integral. This represents the absolute value of the residual. Wherein, and Susceptible to circuit noise pollution, which can be addressed by... Perform Kalman filtering on The process uses a moving average filter, with the filter window approximately equal to 10 × thermal time constant. When the accumulated amount exceeds the threshold, the warning device will fail due to aging. Abnormal oscillations indicate a malfunction in the cooling system.

[0126] Through real-time feature extraction and adaptive compensation closed-loop, this application can maintain sub-microvolt stability under scenarios such as drastic temperature changes and long-term aging, providing core protection for high-precision measurement systems.

[0127] Furthermore, this embodiment employs a mean function and a standard deviation function to jointly construct the framework of the Monte Carlo incremental model. Both the mean function and the standard deviation function are functions of the temperature gradient ∇T. The mean function is based on the inherent drift coefficient and residual temperature drift coefficient derived from the hardware design, while the standard deviation function is determined by the initial random perturbation and the gain term that increases with aging, and is updated over the sampling period. These parameters collectively constitute the model parameters of the Monte Carlo incremental model.

[0128] The expression for the Monte Carlo incremental model is:

[0129]

[0130] in, Represents the mean function; Represents the standard deviation function; Represents a random number between 0 and 1.

[0131]

[0132] in, This represents the temperature compensation coefficient derived from the hardware design. This represents the residual temperature drift coefficient.

[0133]

[0134] in, This represents the initial random perturbation. Indicates the aging coefficient. This indicates the temperature disturbance gain.

[0135] Optionally, when the mean function includes a temperature variation compensation coefficient and the standard deviation function includes an aging coefficient and a temperature perturbation gain, the method for calibrating the Monte Carlo incremental model also includes:

[0136] The temperature change compensation coefficient was obtained by measuring the linear regression relationship between the residual and the square of the rate of temperature change under rapid temperature cycling conditions.

[0137] An aging test was conducted at a constant temperature change rate to establish a linear relationship between the running time and the measured standard deviation function, and the aging coefficient was obtained.

[0138] The standard deviation of the residuals is measured at different rates of temperature change, and the temperature disturbance gain is determined based on the slope of the residuals as a function of the rate of temperature change.

[0139] Specifically, in addition to providing the above functional equations, calibrating the Monte Carlo incremental model also requires determining the parameters of each function, thereby obtaining a definite Monte Carlo incremental model.

[0140] For example, in the temperature change compensation coefficient During calibration, rapid temperature cycling is performed, for example, at a rate of 5–20 °C / min. The linear regression relationship between the residual and the square of the rate of temperature change is then measured, and its expression is as follows:

[0141]

[0142] Where cov(·,·) represents covariance; var(·) represents variance.

[0143] In terms of temperature perturbation gain During calibration, the standard deviation of the residuals is measured at different temperature change rates, and its expression is:

[0144]

[0145] The above scheme employs a rapid temperature cycling method to obtain a temperature change compensation coefficient, ensuring timely correction of the mean function with the rate of temperature change. An aging test establishes a linear relationship between operating time and the standard deviation function, yielding an aging coefficient that makes the fluctuation intensity of the standard deviation function sensitive to aging without drifting. Residual standard deviations are measured at different temperature change rates to determine the temperature disturbance gain, enabling the variance characterization to adapt to scenarios of gradual or rapid temperature changes. Baseline offset and systematic drift terms are determined under statically stable conditions. The resulting Monte Carlo incremental model ensures precise matching of the random incremental statistics for each sampling step with the current temperature change, and the Monte Carlo sample trajectory covers true uncertainty while avoiding divergence. Compared to traditional static compensation, it can operate long-term while maintaining high accuracy and robustness.

[0146] Please see Figure 6 Optionally, a pre-calibrated Monte Carlo incremental model is used to calculate the residual drift value of each sample trajectory at the current step size based on the measurement data, including:

[0147] Step 602: Using the Monte Carlo incremental model, calculate the mean and standard deviation of the measurement data corresponding to the current step length; introduce random perturbation based on the mean and standard deviation to obtain the residual drift increment of the current step length.

[0148] Step 604: Add the residual drift increment of each sample trajectory at the current step length to the residual drift value of the previous step length to obtain the residual drift value of each sample trajectory at the current step length.

[0149] Specifically, at each preset step size, measurement data is acquired once, and the measurement data is input into the aforementioned mean function and standard deviation function to calculate the corresponding mean and standard deviation values. Based on this, a random number Zi is randomly assigned to each sample trajectory, thereby obtaining multiple different incremental samples.

[0150] For the i-th sample trajectory, the expression for its residual drift increment is:

[0151]

[0152] in, Indicates the discrete time step. Used to ensure the invariance of variance scale.

[0153] The residual drift increment is added to the residual drift value calculated in the previous step length to obtain the residual drift increment for the current step length, which is expressed as follows:

[0154]

[0155] Repeat the above process N times until the calculation of all samples is completed, and obtain the result. The sample set represents the cluster of residual temperature drift trajectories that may appear under the current characteristic conditions.

[0156] Optionally, the true output data is reconstructed based on the residual drift values ​​at each current step, resulting in a candidate output set, including:

[0157] For each sample trajectory, the output voltage of the target element is subtracted from the residual drift value of the sample trajectory at the current step size to obtain the true output data; each true output data constitutes a candidate output set.

[0158] Specifically, for the i-th sample trajectory, a virtual signal is constructed:

[0159]

[0160] The actual output voltage is subtracted from the residual drift value to generate N possible predicted distributions. These predicted distributions are free from the influence of temperature drift, thus obtaining the true paths that the output voltage may have.

[0161] Optionally, the number of the above sample trajectories is dynamically set, and the number is positively correlated with the absolute value of the temperature change rate in a hyperbolic tangent relationship.

[0162] Specifically, the expression for the mapping relationship between the number of sample trajectories and the rate of temperature change is:

[0163]

[0164] in, Represents the minimum number of samples in steady state. This represents the maximum number of samples taken during drastic temperature changes. This is the sampling density growth coefficient.

[0165] By adopting the above scheme, the number of sample trajectories is adjusted in real time according to the rate of temperature change, and is constrained by the minimum and maximum number of samples. This reduces computational power consumption and improves the inference speed of the model in scenarios with gradual temperature changes, thus meeting real-time requirements. Simultaneously, the number of sample trajectories automatically matches the temperature-changing scenario, avoiding overcompensation, improving interval reliability, and further enhancing inference accuracy.

[0166] Please see Figure 7 Optionally, a preset decision-making strategy is used to fuse the candidate output set to obtain prediction information, including:

[0167] Step 702: Based on statistical distance metrics, evaluate the consistency between each candidate value and its corresponding sample trajectory in the candidate output set and the set distribution, identify and exclude abnormal trajectories and candidate values ​​whose corresponding candidate outputs deviate significantly from the population distribution, and obtain an effective candidate set.

[0168] Step 704: Based on the similarity between the current temperature change rate and the temperature change rate of each sample trajectory, assign weights to the effective candidate set to obtain a weighted candidate set.

[0169] Step 706: Perform quantile aggregation on the weighted candidate set to generate point estimates and confidence intervals for the current step size, and obtain prediction information.

[0170] Specifically, in order to extract the most reliable prediction information from the large number of random trajectories output by the Monte Carlo incremental model, this application combines three intelligent mechanisms—anomaly removal, trajectory weighting, and quantile aggregation—to further fuse the candidate output set.

[0171] Anomaly removal strategy is used to identify and exclude abnormal trajectories that deviate from the group based on statistical distance metrics, thereby obtaining an effective candidate set.

[0172] A trajectory weighting strategy is used to assign weights to each trajectory in the valid candidate set, prioritizing the trajectory most similar to the current temperature dynamic. The weights of each trajectory have a negative exponential relationship with the Euclidean distance between the current temperature change rate and the trajectory temperature change rate, and their weight expressions are as follows:

[0173]

[0174] in, This represents the current measured rate of temperature change. Let λ represent the rate of temperature change of trajectory i, and λ be the weight decay coefficient.

[0175] Quantile aggregation strategy is used to fuse the trajectories of various samples to obtain optimal prediction information. Predicted values ​​at specific probability quantiles are selected and weighted and fused according to a weight inversely proportional to the probability product. The expression is as follows:

[0176]

[0177]

[0178] Among them, represents the point estimate obtained by fusion; represents the candidate output set; represents the q-th quantile; represents the weight, which is inversely proportional to the product of probabilities and is used to appropriately strengthen the two ends of the quantiles.

[0179] Furthermore, select the upper and lower credible quantiles p1 and p2, where p1 < p2; give the credible interval of the current step size:

[0180]

[0181] Adopting the above scheme, the point estimate is used to calculate and apply the temperature compensation amount within the current step size, and the credible interval is used to evaluate the estimation uncertainty and perform adaptive gain adjustment on the temperature compensation amount to suppress over-compensation or under-compensation and improve the compensation accuracy. The above decision-making strategy converts probability prediction into an executable decision, enabling the temperature drift compensation system to have the industrial intelligent characteristics of self-evaluation, self-optimization, and self-warning, providing core guarantee for high-reliability electronic systems.

[0182] Optionally, generating a temperature drift compensation amount based on prediction information includes:

[0183] Generating an initial compensation amount for the current step size according to the point estimate; based on the credible interval, performing adaptive processing on the initial compensation amount to obtain the temperature drift compensation amount.

[0184] Specifically, determine the direction and magnitude of compensation according to the point estimate to form an initial compensation amount; according to the confidence level of the credible interval, determine whether compensation is required for the current step size and the size of the compensation, and finally obtain the temperature drift compensation amount. The temperature drift compensation amount is written into the compensation digital-to-analog converter and then converted into an analog current signal, which is injected into the input end of the compensation operational amplifier U2 to cancel the residual temperature drift in the output voltage of the target component.

[0185] Optionally, in an exemplary embodiment, the above temperature compensation method further includes:

[0186] Constructing non-linear characteristic parameters based on temperature data; among them, the non-linear characteristic parameters include at least one of the temperature change rate, the change amount of the temperature change rate, and the time decay term of the temperature change intensity;

[0187] Performing weighted summation on each non-linear characteristic parameter based on an adaptive weight to obtain a fine-tuning compensation amount;

[0188] Based on the fine-tuning compensation amount, performing feed-forward correction on the initially compensated amount that has undergone adaptive processing to obtain the temperature drift compensation amount.

[0189] Specifically, the above-mentioned fine-tuning compensation amount is calculated based on a nonlinear compensation model, and its expression is as follows:

[0190]

[0191] in, Indicates the amount of fine-tuning compensation; Indicates the rate of temperature change; This indicates the change in the rate of temperature change; Indicates the intensity of temperature change; This indicates exponentially decaying memory, which characterizes how the short-term inertia or aging effect after a mutation decays over time. It is a time constant; , , This represents adaptive weights.

[0192] At each preset step, the error between the estimate of the residual drift of the nonlinear compensation model and the estimate of the residual drift of the Monte Carlo incremental model is compared. Based on the error and weighted by confidence, a small step update is performed, and the above adaptive weights and time constants are adjusted. The adjusted adaptive weights and time constants will be used in the calculation of the next preset step.

[0193] The above approach uses the results of the Monte Carlo incremental model as the main compensation quantity, and then superimposes them with fine-tuning compensation quantities to reduce residuals and jitter, thereby improving prediction accuracy.

[0194] Optionally, in an exemplary embodiment, the temperature compensation method described above further includes:

[0195] Calibration instructions are generated based on the predicted information to adjust the parameters of the hardware compensation unit; and / or the parameters of the Monte Carlo incremental model are adjusted based on the temperature compensation effect of the target component.

[0196] Specifically, the embodiments of this application are based on a two-way coupling mechanism between the hardware and software layers. Through deep collaboration between hardware and software, temperature drift compensation is achieved by moving from traditional single-level suppression to multi-level collaborative optimization, thereby realizing high-precision measurement.

[0197] The forward learning process refers to the probabilistic model generating hardware calibration instructions to adjust the parameters of the hardware compensation unit. For example, the parameters of the hardware compensation unit include compensation coefficients. Then, a calibration instruction is generated based on the prediction information to adjust the parameters of the hardware compensation unit, including: estimating the slope of the residual drift with respect to temperature. If the slope is greater than 0, it indicates that the output increases with temperature and the compensation coefficient needs to be reduced; otherwise, it increases so as to fine-tune the compensation coefficient accordingly.

[0198] The reverse adaptation process involves detecting hardware parameter drift, adjusting the probabilistic model parameters, and then adjusting the probabilistic model in reverse to compensate for hardware drift.

[0199] Optionally, in an exemplary embodiment, the temperature compensation method described above further includes:

[0200] Based on temperature data, calculate the temperature change rate; compare the temperature change rate with a preset threshold, and switch the working mode according to the comparison result; the working modes include energy-saving mode, high-energy mode, and standard mode.

[0201] Specifically, the energy-saving mode is suitable when the ambient temperature changes little. The activation conditions are when the temperature change rate is lower than the preset threshold and the duration reaches the set value, or when the system idle time reaches the set value.

[0202] The high-energy mode is suitable for rapid temperature changes. In this mode, the system increases the sampling rate, enables high-frequency algorithms, and operates at ultra-high frequency.

[0203] Standard mode is the default startup mode. The condition for switching from energy-saving mode is that the temperature gradient is greater than the preset threshold; the condition for switching from high-energy mode is that the temperature change rate reaches the threshold and the duration reaches the set value.

[0204] To achieve intelligent switching between energy-saving mode, standard mode and high-energy mode, an intelligent switching controller is introduced. It intelligently switches according to threshold switching conditions. Through bidirectional learning and intelligent mode switching, the temperature drift compensation system achieves dynamic optimization in multiple dimensions such as accuracy, power consumption and response speed, providing core capabilities for various electronic systems to adapt to environmental changes.

[0205] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0206] Based on the same inventive concept, this application also provides a temperature compensation system. This system is applicable to the above-described temperature compensation method. The solution provided by this system is similar to the solution described in the above-described method. Therefore, the specific limitations of one or more device embodiments provided below can be found in the limitations of the method above, and will not be repeated here.

[0207] Please see Figure 8 In one embodiment, the temperature compensation system includes: an acquisition module, a prediction module, and a compensation module.

[0208] The acquisition module is used to acquire measurement data of the target component at preset intervals; the measurement data includes temperature data and output voltage after hardware compensation.

[0209] The prediction module uses a pre-calibrated Monte Carlo incremental model to calculate the residual drift value of each sample trajectory at the current step size based on the measurement data. Each sample trajectory is generated by the Monte Carlo incremental model based on the measurement data. The real output data is reconstructed based on the residual drift value at each current step size to obtain a candidate output set. The candidate output set is fused using a preset decision strategy to obtain prediction information.

[0210] The compensation module is used to generate temperature drift compensation based on the predicted information, so that the hardware compensation unit can perform temperature compensation on the target component according to the temperature drift compensation.

[0211] Optionally, the temperature compensation system described above also includes a calibration module.

[0212] The calibration module is used to calibrate the Monte Carlo incremental model. The calibration method includes: acquiring sampled data of the target component at preset sampling intervals. The sampled data includes temperature data, output voltage, and reference voltage; the output voltage is obtained after hardware compensation of the target component; calculating the residual drift voltage based on the output voltage and reference voltage; constructing temperature drift characteristic parameters based on the residual drift voltage and temperature data; the temperature drift characteristic parameters include at least one of the following: temperature change rate, rate of change of temperature change, residual integral, and absolute value of the residual; and constructing the mean function and standard deviation function of the residual drift increment based on the temperature drift characteristic parameters to obtain the Monte Carlo incremental model.

[0213] Optionally, when the mean function includes a temperature change compensation coefficient and the standard deviation function includes an aging coefficient and a temperature disturbance gain, the calibration module calibrates the Monte Carlo incremental model by further including: measuring the linear regression relationship between the residual and the square of the temperature change rate under rapid temperature cycling conditions to obtain the temperature change compensation coefficient; conducting aging tests at a constant temperature change rate to establish a linear relationship between the running time and the measured standard deviation function to obtain the aging coefficient; and measuring the standard deviation of the residual at different temperature change rates and determining the temperature disturbance gain based on its slope with respect to the amplitude of the temperature change rate.

[0214] Optionally, the prediction module employs a pre-calibrated Monte Carlo incremental model to calculate the residual drift value of each sample trajectory at the current step length based on the measurement data. This includes: using the Monte Carlo incremental model to calculate the mean and standard deviation of the measurement data at the current step length; introducing random perturbation based on the mean and standard deviation to obtain the residual drift increment at the current step length; and superimposing the residual drift increment of each sample trajectory at the current step length onto the residual drift value of the previous step length to obtain the residual drift value of each sample trajectory at the current step length. The number of sample trajectories is dynamically set, and the number exhibits a hyperbolic tangent relationship positively correlated with the absolute value of the temperature change rate.

[0215] Optionally, the prediction module reconstructs the true output data based on the residual drift value of each current step to obtain a candidate output set, including: for each sample trajectory, subtracting the residual drift value of the sample trajectory at the current step from the output voltage of the target element to obtain the true output data; each true output data constitutes a candidate output set.

[0216] Optionally, the prediction module uses a preset decision strategy to fuse the candidate output set to obtain prediction information, including: evaluating the consistency between each candidate value and its corresponding sample trajectory and the set distribution based on statistical distance metrics, identifying and excluding abnormal trajectories and candidate values ​​whose corresponding candidate outputs significantly deviate from the population distribution, and obtaining a valid candidate set; assigning weights to the valid candidate set based on the similarity between the current temperature change rate and the temperature change rate of each sample trajectory, and obtaining a weighted candidate set; performing quantile aggregation on the weighted candidate set to generate a point estimate and confidence interval for the current step size, and obtaining prediction information.

[0217] Optionally, the compensation module generates temperature drift compensation based on the prediction information, including: generating an initial compensation amount for the current step size based on point estimation; and adaptively processing the initial compensation amount based on the confidence interval to obtain the temperature drift compensation amount.

[0218] Optionally, the temperature compensation system described above also includes a nonlinear compensation module.

[0219] The nonlinear compensation module is used to construct nonlinear characteristic parameters based on temperature data. The nonlinear characteristic parameters include at least one of the following: temperature change rate, change in temperature change rate, and time decay term of temperature change intensity. Based on adaptive weights, the nonlinear characteristic parameters are weighted and summed to obtain the fine-tuning compensation amount. Based on the fine-tuning compensation amount, the initial compensation amount after adaptive processing is feedforward corrected to obtain the temperature drift compensation amount.

[0220] Optionally, the temperature compensation system described above also includes a calibration module.

[0221] The calibration module is used to generate calibration instructions based on the predicted information to adjust the parameters of the hardware compensation unit; and / or adjust the parameters of the Monte Carlo incremental model based on the temperature compensation effect of the target component.

[0222] Optionally, the temperature compensation system described above also includes a working mode switching module.

[0223] The working mode switching module is used to calculate the temperature change rate based on temperature data; compare the temperature change rate with a preset threshold, and switch the working mode according to the comparison result; the working modes include energy-saving mode, high-energy mode, and standard mode.

[0224] The aforementioned temperature compensation system acquires measurement data of the target component after hardware compensation; processes the measurement data using a pre-calibrated Monte Carlo incremental model to calculate the residual drift value of each sample trajectory at the current step size; reconstructs the true output data based on each residual drift value to obtain a candidate output set; fuses the candidate output set using a preset decision strategy to obtain prediction information; and generates a temperature drift compensation amount based on this prediction information to compensate the temperature of the target component. This application first uses hardware compensation to eliminate most of the linear temperature drift. Based on this, it further employs a Monte Carlo incremental model to accurately predict the temperature drift trajectory at each step size, and then compensates the target component in real time according to the predicted temperature drift trajectory, eliminating the remaining small portion of nonlinear temperature drift, improving the temperature compensation effect of the target component, and achieving high-precision measurement under wide temperature range variations.

[0225] The modules in the aforementioned temperature compensation system can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in the processor of a computer device in hardware form or independent of it, or stored in the memory of the computer device in software form, so that the processor can call and execute the corresponding operations of each module.

[0226] In one feasible embodiment, a computer device is provided, which may be a terminal, and its internal structure diagram may be as follows: Figure 9As shown, the computer device includes a processor, memory, input / output interface, communication interface, display unit, and input device. The processor, memory, and input / output interface are connected via a system bus, and the communication interface, display unit, and input device are also connected to the system bus via the input / output interface. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The input / output interface is used for exchanging information between the processor and external devices. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, mobile cellular networks, NFC (Near Field Communication), or other technologies. When the computer program is executed by the processor, it implements the aforementioned temperature compensation method. The display unit is used to form a visually visible image and can be a display screen, a projection device, or a virtual reality imaging device. The display screen can be an LCD screen or an e-ink screen. The input device of the computer device can be a touch layer covering the display screen, or buttons, trackballs, or touchpads set on the casing of the computer device, or external keyboards, touchpads, or mice, etc.

[0227] Those skilled in the art will understand that Figure 9 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0228] In one feasible embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the method steps in the temperature compensation method described above.

[0229] In one feasible embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the method steps in the temperature compensation method described above.

[0230] In one feasible embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the method steps in the temperature compensation method described above.

[0231] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0232] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0233] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A temperature compensation method, characterized in that, The method includes: Measurement data of the target component is acquired at preset intervals; wherein, the measurement data includes temperature data and output voltage after compensation by a hardware compensation unit; A pre-calibrated Monte Carlo incremental model is used to calculate the residual drift value of each sample trajectory at the current step size based on the measurement data; wherein, each sample trajectory is generated by the Monte Carlo incremental model based on the measurement data; Based on the residual drift values ​​of each current step, the true output data is reconstructed to obtain a candidate output set; The candidate output set is fused using a preset decision-making strategy to obtain prediction information; Based on the predicted information, a temperature drift compensation amount is generated so that the hardware compensation unit can perform temperature compensation on the target component according to the temperature drift compensation amount.

2. The method according to claim 1, characterized in that, The hardware compensation unit includes two symmetrically designed operational amplifiers, and the temperature drift characteristics of the two operational amplifiers are matched.

3. The method according to claim 1, characterized in that, Methods for calibrating Monte Carlo incremental models include: Sampling data of the target component is acquired at preset sampling intervals, wherein the sampling data includes temperature data, output voltage, and reference voltage; the output voltage is obtained after hardware compensation of the target component. The residual drift voltage is calculated based on the output voltage and the reference voltage. Based on the residual drift voltage and the temperature data, temperature drift characteristic parameters are constructed; wherein, the temperature drift characteristic parameters include at least one of the following: temperature change rate, rate of change of temperature change rate, residual integral, and residual absolute value; Based on the temperature drift characteristic parameters, the mean function and standard deviation function of the residual drift increment are constructed to obtain the Monte Carlo incremental model.

4. The method according to claim 3, characterized in that, The mean function includes a temperature change compensation coefficient; the standard deviation function includes an aging coefficient and a temperature disturbance gain. Methods for calibrating Monte Carlo incremental models also include: The temperature change compensation coefficient is obtained by measuring the linear regression relationship between the residual and the square of the rate of temperature change under rapid temperature cycling conditions. An aging test was conducted at a constant temperature change rate, and a linear relationship was established between the running time and the measured standard deviation function to obtain the aging coefficient. The standard deviation of the residuals is measured at different rates of temperature change, and the temperature disturbance gain is determined based on the slope of the residuals as a function of the rate of temperature change.

5. The method according to claim 1, characterized in that, The method employs a pre-calibrated Monte Carlo incremental model to calculate the residual drift value of each sample trajectory at the current step size based on the measurement data, including: Using the Monte Carlo incremental model, the mean and standard deviation of the measurement data for the current step size are calculated; random perturbation is introduced based on the mean and standard deviation to obtain the residual drift increment for the current step size; The residual drift increment of each sample trajectory at the current step length is superimposed on the residual drift value of the previous step length to obtain the residual drift value of each sample trajectory at the current step length.

6. The method according to claim 5, characterized in that, The number of sample trajectories is dynamically set, and the number is positively correlated with the absolute value of the temperature change rate in a hyperbolic tangent relationship.

7. The method according to claim 1, characterized in that, The process of reconstructing the true output data based on the residual drift values ​​of each current step size yields a candidate output set, including: For each sample trajectory, the output voltage of the target component is subtracted from the residual drift value of the sample trajectory at the current step size to obtain the true output data; The actual output data constitute the candidate output set.

8. The method according to claim 1, characterized in that, The step of fusing the candidate output set using a preset decision-making strategy to obtain prediction information includes: Based on statistical distance metrics, the consistency between each candidate value and its corresponding sample trajectory in the candidate output set and the set distribution is evaluated. Abnormal trajectories and candidate values ​​whose corresponding candidate outputs deviate significantly from the population distribution are identified and excluded to obtain an effective candidate set. Based on the similarity between the current temperature change rate and the temperature change rate of each sample trajectory, weights are assigned to the effective candidate set to obtain a weighted candidate set. Quantile aggregation is performed on the weighted candidate set to generate a point estimate and confidence interval for the current step size, thus obtaining the prediction information.

9. The method according to claim 8, characterized in that, Generating temperature drift compensation based on the predicted information includes: The initial compensation amount for the current step size is generated based on the point estimate; Based on the confidence interval, the initial compensation amount is adaptively processed to obtain the temperature drift compensation amount.

10. The method according to claim 9, characterized in that, The method further includes: Based on the temperature data, nonlinear characteristic parameters are constructed; wherein, the nonlinear characteristic parameters include at least one of the following: temperature change rate, change in temperature change rate, and time decay term of temperature change intensity. Based on adaptive weights, the nonlinear characteristic parameters are weighted and summed to obtain the fine-tuning compensation amount. Based on the fine-tuning compensation amount, the initial compensation amount after adaptive processing is feedforward corrected to obtain the temperature drift compensation amount.

11. The method according to claim 1, characterized in that, The method further includes: A calibration instruction is generated based on the predicted information to adjust the parameters of the hardware compensation unit; And / or adjust the parameters of the Monte Carlo incremental model based on the temperature compensation effect of the target element.

12. The method according to claim 1, characterized in that, The method further includes: Calculate the rate of temperature change based on the temperature data; The temperature change rate is compared with a preset threshold, and the working mode is switched according to the comparison result; wherein, the working mode includes energy-saving mode, high-energy mode and standard mode.

13. A temperature compensation system, characterized in that, The system includes: The acquisition module is used to acquire measurement data of the target component at preset intervals; wherein, the measurement data includes temperature data and output voltage after compensation by a hardware compensation unit; The prediction module employs a pre-calibrated Monte Carlo incremental model to calculate the residual drift value of each sample trajectory at the current step length based on the measurement data. Each sample trajectory is generated by the Monte Carlo incremental model based on the measurement data. The module reconstructs the true output data based on the residual drift values ​​at each current step length to obtain a candidate output set. A preset decision strategy is then used to fuse the candidate output set to obtain prediction information. The compensation module is used to generate a temperature drift compensation amount based on the predicted information, so that the hardware compensation unit can perform temperature compensation on the target component according to the temperature drift compensation amount.

14. A testing device, characterized in that, include: The data acquisition module is used to acquire temperature information of the wafer during the thin film epitaxial growth process. A temperature detection module is used to monitor the temperature change of the wafer during the thin film growth process based on the temperature information; the temperature detection module includes a target element and a hardware compensation unit. The temperature compensation system as described in claim 13 is used to acquire measurement data of the target component and generate a temperature drift compensation amount based on the measurement data, so that the hardware compensation unit performs temperature compensation on the target component according to the temperature drift compensation amount.

15. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1-12.

16. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1-12.

Images