Method and system capable of automatically controlling and supervising chip functions

Abstract

The invention provides a method and system capable of automatically controlling and supervising chip functions, relates to the field of machine vision detection, and is particularly suitable for functional detection and active protection of a chip module in a complex application environment. The method for automatically controlling and supervising the chip function specifically comprises the following steps of: placing a chip module in an incubator and connecting a test circuit; based on the position information of each element in the chip module, the incubator is guided to carry out local temperature rise, and a temperature-position influence infrared image set of all the elements is obtained; starting detection, and obtaining an original infrared time sequence image sequence of the chip module; based on the temperature-position influence infrared image set of each element position and the environment temperature in the original infrared time sequence image sequence, correcting the original infrared time sequence image sequence to obtain a self-heat-generation infrared time sequence image sequence of the chip module; and performing heat production supervision on elements in the self heat production infrared time sequence image sequence based on the time sequence model. Through the steps, the functional change of the chip element is accurately reflected, the interference of the environment and other elements is effectively stripped, and the accuracy of chip health monitoring is greatly improved.

Description

Technical Field

[0001] The present invention relates to the field of machine vision inspection, and in particular to a method and system for automatically controlling and supervising chip functions, which is particularly suitable for functional inspection and active protection of chip modules in complex application environments. Background Art

[0002] With the rapid development of artificial intelligence (AI) technology, smart chips have been widely used in a variety of high-performance and high-security scenarios, including cloud computing, autonomous driving, and smart terminals. In modern industrial production, functional testing and temperature monitoring of chips have become critical steps in ensuring system safety and stable operation. After prolonged operation, the internal semiconductor structure, nodes, and other key components of a chip undergo irreversible aging, resulting in significant deviations between actual operating parameters (such as energy consumption, heat generation, and operating current) and factory calibration values. If the chip is still mechanically managed and scheduled according to factory calibration parameters during subsequent use, it can easily lead to significant safety risks such as overheating, performance degradation, and even failure. Therefore, how to dynamically and accurately monitor the actual operating status of a chip throughout its lifecycle and adjust management measures accordingly has become a pressing technical challenge for the industry.

[0003] Existing chip testing technologies primarily fall into two categories. The first involves discrete testing of the functionality and performance indicators of individual chip components. This type of testing is often conducted in idealized laboratory environments, effectively eliminating interference from the external environment and other components, and determining the performance limits of the chip component itself. However, in real-world applications, chips often operate in modules or integrated circuit boards, subject to complex and variable environmental conditions (such as temperature, humidity, and electromagnetic interference), and interact with other chips and components in various ways, including energy, heat, and noise. In these situations, test data obtained under ideal conditions often fails to truly reflect the functional performance of chip components under actual operating conditions. Semiconductor materials' heat generation characteristics, aging rates, and heat dissipation efficiency vary depending on environmental conditions, making laboratory test results of limited value in practical applications. The second approach involves functional testing of the entire chip module, specifically testing a complete module that integrates multiple different types of components. Due to the coupling and interactions between components within the module, the temperature rise of a single chip component is affected not only by its own characteristics but also by the heat generated by nearby components and changes in the external environment. Existing technologies struggle to effectively isolate the impact of these complex factors on target chip components, resulting in significant variability in test results and an inability to accurately reflect the true functionality and health of the target components themselves. Therefore, whether testing discrete components or entire modules, significant issues exist, including inconsistencies between the test environment and the actual operating environment, complex and difficult-to-decouple influencing factors, and insufficient reliability and specificity of test results. These issues directly impact the accuracy of chip functionality assessments and the scientific nature of subsequent management measures, posing a threat to the safe operation of chips throughout their lifecycle.

[0004] In response to the above-mentioned deficiencies in the prior art, the present invention proposes a method and system that can automatically control and monitor chip functions. The system obtains the thermal distribution information of the chip module under real working conditions based on the infrared image sequence, and combines it with standardized test conditions to obtain the actual heat generation of the semiconductor material inside the chip. By introducing a timing network model, the dynamic thermal features in the infrared image sequence are extracted and analyzed to achieve real-time prediction of changes in the chip's own working conditions. The system can actively predict risks such as abnormal temperature rise in chip functions, dynamically adjust chip operating parameters or issue protection instructions, thereby achieving intelligent management and active protection throughout the chip's life cycle. Compared with existing detection methods, the present invention can accurately reflect the functional changes of the chip components themselves without affecting the normal operation of the chip, effectively remove interference from the environment and other components, and greatly improve the accuracy of chip health monitoring and the safety of engineering applications. Summary of the Invention

[0005] In view of this, the present invention provides a method for automatically controlling and supervising chip functions, the method specifically comprising the following steps: S1: Place the chip module in a temperature chamber and connect the test circuits; S2: Based on the position information of each component in the chip module, the temperature box is guided to perform local temperature increase, and a set of infrared images of the temperature-position effect of all components is obtained; S3: Start detection and obtain the original infrared time-series image sequence of the chip module; S4: Based on the temperature-position influence infrared atlas of each component position and the ambient temperature in the original infrared time-series image sequence, the original infrared time-series image sequence is corrected to obtain the chip module's own heat generation infrared time-series image sequence; S5: Supervise the heat generation of components in the self-heating infrared time-series image sequence based on the time-series model.

[0006] The present invention also provides a system for automatically controlling and supervising chip functions, the system comprising: Temperature box: Place the chip module in the temperature box and connect the test circuits; Local temperature rise control module: Based on the position information of each component in the chip module, it instructs the temperature chamber to perform local temperature rise, and obtains the infrared image set of the temperature-position influence of all components; Infrared time-series image sequence acquisition module: Start detection and use the infrared thermal imager to obtain the original infrared time-series image sequence of the chip module; Image correction module: Based on the temperature-position influence infrared atlas of each component position and the ambient temperature in the original infrared time-series image sequence, the original infrared time-series image sequence is corrected to obtain the chip module's own heat generation infrared time-series image sequence; Supervision module: Based on the time series model, the heat generation of the components in the self-heating infrared time series image sequence is supervised.

[0007] The present invention also provides an electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the method for automatically controlling and supervising chip functions is implemented.

[0008] The present invention also provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method for automatically controlling and supervising chip functions is implemented.

[0009] Compared with the prior art, the present invention discloses a method and system that can automatically control and monitor chip functions. First, the real heat generation contribution of each independent component inside the chip module is accurately decoupled and quantified. Then, based on this self-heating data, a timing and correlation analysis model is used to achieve high-precision, intelligent supervision of the operating status of each component and early abnormality warning. This method overcomes the fundamental problem in traditional chip thermal management and fault diagnosis that it is difficult to distinguish between the self-heating of a component and the thermal influence of the environment and neighboring components. Specifically, the present invention can extract the net heat generated by each component due to its own workload from the complex and mixed original thermal map through a double-cycle temperature-position impact infrared atlas calibration and a triple-cycle correction algorithm. This enables the control system to clearly and accurately understand which components are actually working at high load and which components may have abnormal heat generation, rather than just passive heating. Based on the self-heating infrared time-series image sequence and time-series model, it is not only possible to detect anomalies in the heat generation timing of a single component, but also to analyze the complex correlations between the heat generation patterns of different components. The analysis of these self-level self-timing characteristics and mutual pattern characteristics can reveal chip self-degradation, hardware defects and system-level collaborative work anomalies that are difficult to detect with traditional methods. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0011] Figure 1 It is a flow chart of the chip functions that can be automatically controlled and supervised in this application. DETAILED DESCRIPTION

[0012] The embodiments of the present application are described in detail below with reference to the accompanying drawings.

[0013] The following describes the embodiments of the present application through specific examples, and those skilled in the art can easily understand other advantages and effects of the present application from the contents disclosed in this specification. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. The present application can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present application. It should be noted that, in the absence of conflict, the features in the following embodiments and embodiments can be combined with each other. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without making creative work are within the scope of protection of this application.

[0014] It should be noted that various aspects of the embodiments within the scope of the appended claims are described below. It should be apparent that the aspects described herein can be embodied in a wide variety of forms, and any specific structure and / or function described herein is merely illustrative. Based on this application, it should be understood by those skilled in the art that an aspect described herein can be implemented independently of any other aspect, and two or more of these aspects can be combined in various ways. For example, any number and aspect described herein can be used to implement an apparatus and / or practice a method. In addition, other structures and / or functionalities other than one or more of the aspects described herein can be used to implement this apparatus and / or practice this method.

[0015] Additionally, in the following description, specific details are provided to provide a thorough understanding of the examples, however, one skilled in the art will appreciate that the examples can be practiced without these specific details.

[0016] The following describes the technical solutions provided by various embodiments of the present application in conjunction with the accompanying drawings.

[0017] The embodiments of this specification provide a method for automatically controlling and supervising chip functions, which specifically includes the following steps: S1: Place the chip module in a temperature chamber and connect the test circuits; S2: Based on the position information of each component in the chip module, the temperature box is guided to perform local temperature increase, and a set of infrared images of the temperature-position effect of all components is obtained; S3: Start detection and obtain the original infrared time-series image sequence of the chip module; S4: Based on the temperature-position influence infrared atlas of each component position and the ambient temperature in the original infrared time-series image sequence, the original infrared time-series image sequence is corrected to obtain the chip module's own heat generation infrared time-series image sequence; S5: Supervise the heat generation of components in the self-heating infrared time-series image sequence based on the time-series model.

[0018] In this embodiment, the incubator is an integrated, high-precision environmental control device specifically designed for local and overall temperature regulation and infrared imaging detection of chip modules. The main body of the incubator utilizes a double-layer thermal insulation structure. The outer shell can be made of high-strength engineering plastics or metal materials, while the inner liner is constructed from materials with low thermal conductivity and high infrared reflectivity, such as polished stainless steel or coated alloys. This minimizes internal and external heat exchange and ensures uniformity and stability of the temperature environment within the incubator. A chip module carrier platform is provided within the incubator. The carrier platform is preferably constructed from an insulating material with a low thermal expansion coefficient, such as ceramic or a special composite material. The carrier platform is integrated with a fixture for mounting the chip module to be tested, and ample wiring space and multiple electrical signal feedthrough interfaces are reserved. All wiring is introduced through a sealed wire duct at the rear of the incubator. The incubator's main control system reserves multiple signal channels that can communicate with an external host computer system, enabling unified management and automated control of the incubator's heating process, camera acquisition, and chip module test status. The system supports remote setting of the overall and local temperature parameters of the incubator, and can record in real time the infrared and visible light image sequences of the chip module under different temperature control conditions.

[0019] To accurately simulate and control the chip module's operating environment, the incubator is equipped with a global temperature control system. This system includes an air circulation fan, evenly distributed resistive heating elements, and a Peltier semiconductor cooling module. At least one temperature sensor monitors the air temperature inside the incubator in real time. This temperature sensor is connected to the incubator's main control system and, using a PID control algorithm, precisely adjusts the power output of the heating elements and the operating status of the cooling module. This allows for rapid setting, stable maintenance, and programmed temperature adjustments within the incubator over a wide temperature range. The incubator's internal temperature adjustment range is preferably set to -40°C to +60°C.

[0020] Directly beneath the chip module platform, an array of independently addressable and controllable micro-local heating units is integrated. These local heating units, preferably micro-ceramic heaters, thin-film resistive heaters, or arrays of small thermoelectric modules, are each connected to a multi-channel precision power control system. This system independently and precisely controls the power output of one or more selected local heating units based on component position information and test requirements, thereby achieving precise, rapid, and independent heating operations at specific locations within the chip module. Component position information can be obtained from the chip design layout or through visible light visual inspection.

[0021] In order to conduct real-time non-contact monitoring of the temperature distribution of the chip module during the test, one or more optical observation windows are provided on the top of the incubator. At least one of the windows is made of a special material with high transmittance in the mid- and far-infrared bands. The infrared window faces the chip module carrier platform, and the infrared thermal imager can capture the complete surface temperature distribution image of the chip module under various working conditions. To further assist in the analysis, preferably, the incubator is also provided with another visible light observation window. In conjunction with the adjustable LED lighting source inside the incubator, the visible light camera can capture the physical morphology image of the chip module for positioning, alignment or auxiliary status judgment. The infrared thermal imager and the visible light camera are fixed to the precision adjustment bracket outside the incubator through a standard interface. The data output interfaces of all camera devices are led out to the outside of the incubator and connected to the image acquisition and processing system.

[0022] The chamber is also equipped with a dry nitrogen purge port to prevent internal frost, an over-temperature protection device, and a door locking mechanism to ensure operational safety and tightness. All controllable components of the entire chamber system, including global temperature control, local heating unit array control, and internal lighting control, can be connected to the main control system through a unified control interface, enabling automated control of the test process and synchronous data collection.

[0023] To further illustrate the structure and operating principle of the localized heating unit in the present invention, a specific embodiment based on a thermocouple module array is provided. In this embodiment, the localized heating unit array does not utilize traditional resistance wire heating. Instead, it leverages the precise, high-spatial-resolution temperature control characteristics of semiconductor cooling elements operating in reverse (i.e., heating) mode to achieve targeted heating of any key component on the chip module.

[0024] The local heating unit array in this embodiment is integrated directly below the chip module carrier platform, and its core is a high-density micro-TEC module matrix. The substrate of the array has a ceramic material with high thermal conductivity, such as aluminum nitride or beryllium oxide. The high thermal conductivity ensures that the waste heat generated by the TEC module during operation can be quickly and evenly transferred to the shared heat dissipation system under the carrier platform, preventing heat accumulation inside the substrate, thereby maintaining the temperature reference stability and service life of the entire heating array. The ceramic substrate is processed, and its side facing the chip module carrier platform, that is, the upper surface, needs to be polished to maximize thermal contact efficiency.

[0025] Micro-TEC modules are mounted in an array on the top surface of a ceramic substrate using thermally conductive silver adhesive. These TEC modules are small, preferably ranging from 3mm x 3mm to 5mm x 5mm, forming a heating matrix with high spatial resolution. For example, a 50mm x 50mm area can be integrated with an array of 10x10, or 100, 5mm x 5mm TEC modules.

[0026] Each TEC module consists of multiple P-type and N-type semiconductor dipole arms connected in series, sandwiched between two ceramic substrates. When DC current is applied, one side of the ceramic absorbs heat while the other side releases heat, depending on the direction of the current. In this invention, we exploit this heating function to independently control the temperature of the hot surface of each TEC module by precisely controlling the direction and magnitude of the current applied to it.

[0027] The hot surface of the TEC module array faces upward, directly toward the bottom of the chip module carrier platform. To eliminate the tiny air gap between the TEC module and the carrier platform, a thin, uniform layer of high-performance thermal interface material is filled between the two. Preferably, a phase-change thermal conductive material is used. This phase-change thermal conductive material transforms from a solid to a semi-fluid state at a specific temperature, filling all microscopic gaps and achieving near-zero interfacial thermal resistance. This ensures that the heat generated by the TEC module is efficiently and losslessly transferred to the carrier platform, ultimately precisely affecting the corresponding position of the chip module under test.

[0028] This embodiment uses a matrix addressing architecture to control the TEC array. Multiple layers of wiring are etched on the ceramic substrate to form row and column drive lines. The two electrodes of each TEC module are connected to a specific row line and column line, respectively. When the central control system needs to activate the TEC module at a specific location, it only needs to apply a drive signal to the corresponding row line and column line. For each selected TEC module, the system does not simply apply a constant DC power, but instead uses high-frequency pulse width modulation for power regulation. The central control system outputs a PWM signal to the drive unit of the corresponding TEC module. By adjusting the duty cycle of the PWM signal, the average power delivered to the TEC module per unit time can be controlled. The higher the duty cycle, the greater the heating power and the higher the target temperature.

[0029] To achieve precise stabilization of the target temperature, this system employs a closed-loop feedback control strategy. Miniaturized temperature sensors are integrated into the TEC module array at regular intervals. Preferably, miniature NTC thermistors or thin-film platinum resistors are spaced every four or nine TEC modules. These sensors, coexisting with the TEC modules on the same substrate, accurately measure the actual temperature of a localized area in real time.

[0030] When heating a component, the control system reads the values from one or more micro-temperature sensors closest to the location. This real-time temperature value is fed into a dedicated PID controller as a process variable. The PID controller's setpoint is the target temperature rise required for the current inner loop. The PID controller calculates the error between the PV and SP in real time and dynamically adjusts the duty cycle of the TEC module at that location until the error is eliminated. This closed-loop control approach overcomes the influence of individual TEC module variations and minor environmental fluctuations, ensuring accurate and stable heating temperatures and achieving thermal steady-state conditions.

[0031] Based on the position information of each component in the chip module, the temperature chamber is guided to perform local temperature increase, including: Number N key components in the chip module and obtain corresponding location information ; Set M discrete local target heating temperature values and overall temperature, wherein the overall temperature is the operating environment temperature of the chip module; A double loop structure is used to traverse N key component positions and M discrete local target heating temperature values to obtain M temperature-position effect infrared images of each component, forming a set of temperature-position effect infrared images of all components; Among them, the temperature-position effect infrared image set of all elements composed of a double loop structure specifically includes: Outer loop: Select the component position to be calibrated , , the central control system is based on Activate the local heating unit at its corresponding position.

[0032] Inner loop: For the currently activated local heating unit, select the target heating temperature value in turn Adjust the power applied to the local heating unit so that the temperature of the corresponding area gradually increases and stabilizes at the target heating temperature value; In the inner cycle, under thermal steady-state conditions, the temperature distribution of the entire chip module is collected when each component is heated to the target temperature value; The inner loop also includes a steady-state judgment process: During the heating process, the infrared thermal imager continuously captures infrared thermal images of the entire chip module at preset time intervals; the system monitors the temperature distribution changes of the entire module in real time, and analyzes the position of the component to be calibrated in the continuous infrared images. The temperature change determines whether thermal steady state has been reached.

[0033] The thermal steady state is set as that within a preset time window, the average temperature change rate of the pixel points in the component position area currently to be calibrated is less than a preset temperature threshold.

[0034] Once thermal steady state is confirmed, the local heating unit is heated to Under the condition of stable heating of the target, the temperature distribution of the current component position area to be calibrated no longer changes significantly, and the central control system triggers the infrared thermal imager to collect and store an infrared image ; The infrared image Records the time when the kth element is heated to The temperature distribution of the entire chip module is ; The inner cycle also includes cooling and starting the recirculation process: Complete an infrared image After the data is collected, the central control system turns off the current local heating unit. When the entire chip module cools down to the operating environment temperature, the inner cycle continues and the same component position Apply the next target ramp temperature value , repeat heating, steady-state judgment, infrared image acquisition and storage until all M target heating temperature values are completed.

[0035] The outer loop also includes: when the element position After all the infrared images of the M target temperature values are collected, the outer loop enters the next iteration and selects the new component position , and repeat all the operations of the inner loop.

[0036] This invention quantifies the thermal impact of chip modules. It actively and controllably heats each key component independently and simultaneously records infrared thermal images of the entire module, thereby constructing a comprehensive thermal impact database—the temperature-position impact infrared atlas. This atlas documents the extent and extent to which any given component's heat impacts all other components and regions within the module, providing a critical calibration basis for subsequently accurately delineating inter-component thermal impacts. The atlas forms the foundation of the correction algorithm. Without this atlas, it would be impossible to accurately calculate and deduct the contribution of adjacent heating components to a component's observed temperature during actual operation. This overcomes the problem of traditional infrared thermal imaging, which only reveals an apparent mixture of temperatures. Furthermore, through independent experiments, this invention enhances understanding of the thermal characteristics of specific chip modules. Chip modules with different packages and layouts exhibit significant variations in internal heat transfer characteristics. Through field measurements, the invention establishes unique thermal impact model parameters for each specific module, avoiding reliance on general, imprecise theoretical models or empirical estimates.

[0037] Start the detection and obtain the original infrared time-series image sequence of the chip module including; The overall ambient temperature inside the incubator is adjusted to the operating ambient temperature through the incubator's global temperature control system. The operating ambient temperature is the room temperature in the chip module's working environment. During the overall ambient temperature adjustment process, the heating device in the incubator works in conjunction with the air circulation fan and uses temperature sensor feedback to ensure that the air temperature in the incubator reaches the operating ambient temperature and remains stable. When the temperature of the incubator stabilizes at the operating environment temperature, the original infrared time-series image sequence of the chip module is acquired; The original infrared time-series image sequence acquisition of the chip module includes: Set the start time, apply power to the chip module through the test system, and execute the predetermined workload, triggering the infrared thermal imager to start continuously capturing infrared thermal images of the chip module surface.

[0038] The workload can be an application program under actual working conditions of the chip module, a standard benchmark suite such as SPEC CPU, Linpack, or a series of preset instruction sequences that simulate user operations.

[0039] As workloads execute, key components within the chip module begin performing calculations and processing data, generating heat as a result of their activity. This heat is transferred to the module surface through the chip's heat dissipation structure, causing the surface temperature distribution to change dynamically.

[0040] The infrared thermal imager generates raw infrared image frames at a set frame rate , i is the image frame number from the start time, each frame They truly reflect the complete temperature distribution on the chip module surface at the exact time point.

[0041] This method records the complete surface temperature field of a chip over time while executing an application or specific test program. This infrared time-series image sequence not only shows the spatial distribution of heat on the chip surface but also reveals how the heat changes over time. This two-dimensional raw data is far richer than traditional single-point temperature sensors.

[0042] After the acquisition is completed, the original infrared time-series image sequence is subjected to a preliminary integrity check, which includes checking for frame loss and time stamp continuity. After confirmation, the original infrared time-series image sequence is stored together with all related metadata. The metadata should at least include: the set temperature of the overall operating environment of the incubator, a detailed description or identification of the workload being performed, the model and acquisition parameters of the infrared thermal imager, the start and end time of the test, and the model of the chip module. Based on the temperature-position effect infrared atlas of each component position and the ambient temperature in the original infrared time-series image sequence, the original infrared time-series image sequence is corrected to obtain the chip module's own heat generation infrared time-series image sequence; A triple loop structure is used to obtain the chip module's self-heating infrared time-series image sequence, where the first loop iteration includes extracting any original infrared image frame in the original infrared time-series image sequence. , generate a corresponding corrected image , the correction process is carried out for N key components. Each critical component in the system needs to be calculated and the thermal impact of all other components on it needs to be removed.

[0043] For each raw infrared image frame The correction process sets a second loop structure, which traverses all key elements to be corrected ; For the currently selected key element , get the key component in Corresponding position in The average observed temperature ; For each key component to be corrected The correction process sets a third loop structure that traverses all other key elements , , ; Correct each key element to be corrected based on the third loop structure include: For other key components , get the key component exist Corresponding position in The average observed temperature , access the temperature-position effect infrared map corresponding to the key components M infrared images, find Corresponding to the position in M infrared images The average temperature of the current frame The average observed temperature The closest one , calculated At temperature Time The resulting temperature rise is: ; in, represents the temperature rise effect of key component j on key component k in the i-th frame, Indicates the target temperature rise value of key component j The average temperature of the key component k position in the obtained temperature-position-affecting infrared image; Operating ambient temperature; The calculated key components For key components Thermal influence Added to the total external heat impact: ; in, is the key element of frame i The total temperature rise affected; The third cycle continues, for the same key component Apply the next other key element until all other key elements are traversed, and obtain the total temperature increase effect of all other elements on the key element k .

[0044] Utilizing key components Location The average observed temperature Get the corrected temperature at that location: ;in, is the position in the i-th frame Corrected temperature; Modify each original infrared image frame based on the second loop structure Also includes: Set the original infrared image frame Each key element in is a key element to be corrected, and the third loop structure is executed until the original infrared image frame Each key component in the image eliminates the thermal impact of other key components and obtains the corresponding corrected image. ; Repeated execution for a single frame The correction process is completed until each frame of the original infrared time-series image sequence has been processed and its corresponding corrected image is generated; all corrected image frames are combined in time sequence, and the chip module's own thermal infrared time-series image sequence is recorded as , is the number of frames in the original infrared time-series image sequence. Each frame in the chip module's self-heating infrared time-series image sequence represents the net heat contribution generated by each key component due to its own operation at a specific moment.

[0045] Through the above-mentioned corrections, the present invention achieves the precise separation and extraction of the heat generated by each component itself. The temperature-position effect infrared image set obtained by S2 is used to correct the original infrared time-series image sequence collected by S3, effectively eliminating the contamination of the observed temperature of the current target component by other heating components through the heat source. The resulting self-heating infrared time-series image sequence can more realistically reflect the net heat generated by each independent component due to its own operating power consumption. In traditional methods, a component with a very high surface temperature may be because it generates a lot of heat itself, or it may be because it is next to a large furnace component. The correction of this step makes it possible to clearly distinguish between these two situations, thereby accurately locking in the component that is actually causing abnormal heating due to its own problems and avoiding misjudgment.

[0046] The heat generation supervision of components in the self-heating infrared time series image sequence based on the time series model includes: Component 2D image block sequence extraction; Detection of abnormal heating in individual component regions based on ConvLSTM autoencoder; Correlation analysis of changes in self-heating patterns among key component regions; Generation and output of heat production supervision information.

[0047] The component two-dimensional image block sequence extraction includes for each key component ,from Each frame of the image Based on its location information Extract the corresponding 2D image block , thus, for each key element Construct its independent 2D image block time series ; The component also includes sliding window segmentation after extracting the two-dimensional image block sequence; defining a time window of fixed length and sliding step length At any time ,element The input data is a sequence of image blocks ; In order to analyze and detect the heat generation behavior of each key component area, this embodiment provides a The image block sequence is configured with an independent convolutional long short-term memory network autoencoder, denoted as . Corresponding components The training is performed using data generated under normal fault-free operation conditions.

[0048] Encoder in Composed of ConvLSTM layers, encoder take over A sequence of image blocks As input, the spatiotemporal features are extracted by stacking multiple layers of ConvLSTM, and the image block sequence is compressed and flattened into a one-dimensional feature vector. , , Decoder in , whose structure is symmetrical with the encoder, it receives a one-dimensional feature vector As input, it is decoded and reconstructed into the original image block sequence . .

[0049] For a sequence of image blocks , input it into the pre-trained In the image block sequence window, we get the reconstructed image block sequence window , calculate the reconstruction error of the current window as its spatiotemporal anomaly score , the anomaly score and the component Preset Compare. If , then the decision element Region in the current time window The self-heating behavior in the body is abnormal in time and space, and an abnormal sign is generated. Otherwise .

[0050] For the current analysis window , calculate any two different components and , , the one-dimensional eigenvector of and The similarity between ,when , is the similarity threshold, then the component is considered and There is a significant correlation between the spatiotemporal patterns of regional heat production within the current window.

[0051] The present invention uses a timing model to process a two-dimensional self-heating infrared timing image sequence, which can learn and identify the spatiotemporal dynamic pattern of self-heating of each component area under normal working conditions. Compared with simple threshold judgment, the timing model can capture subtle temporal changes in the heat generation pattern, such as abnormal heat generation rise rate, irregular heat generation fluctuations, etc. These anomaly detections based on self-timing characteristics can detect potential functional degradation or early signs of failure of components earlier than traditional methods. By analyzing the correlation and synergy between the spatiotemporal patterns of self-heating of different component areas, more complex system-level problems can be discovered in addition to the mutual influence of component heat. Abnormal synchronous or asynchronous changes in the self-heating patterns of multiple components indicate problems with shared resources or chain reactions triggered by specific task flows within the chip. This mutual pattern analysis provides system-level detection beyond the perspective of a single component.

[0052] The present invention also provides a system for automatically controlling and supervising chip functions, the system comprising: Temperature box: Place the chip module in the temperature box and connect the test circuits; Local temperature rise control module: Based on the position information of each component in the chip module, it instructs the temperature chamber to perform local temperature rise, and obtains the infrared image set of the temperature-position influence of all components; Infrared time-series image sequence acquisition module: Start detection and use the infrared thermal imager to obtain the original infrared time-series image sequence of the chip module; Image correction module: Based on the temperature-position influence infrared atlas of each component position and the ambient temperature in the original infrared time-series image sequence, the original infrared time-series image sequence is corrected to obtain the chip module's own heat generation infrared time-series image sequence; Supervision module: Based on the time series model, the heat generation of the components in the self-heating infrared time series image sequence is supervised.

[0053] The present invention also provides an electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the method for automatically controlling and supervising chip functions is implemented.

[0054] The present invention also provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method for automatically controlling and supervising chip functions is implemented.

[0055] Those skilled in the art will appreciate that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0056] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in the present application should be included in the scope of protection of the present application. Therefore, the scope of protection of the present application should be based on the scope of protection of the claims.

Claims

1. A method for automatically controlling and supervising chip functions, characterized in that: The method comprises the following steps: S1: Place the chip module in a temperature chamber and connect the test circuits; S2: Based on the position information of each component in the chip module, the temperature box is guided to perform local temperature increase, and a set of infrared images of the temperature-position effect of all components is obtained; S3: Start detection and obtain the original infrared time-series image sequence of the chip module; S4: Based on the temperature-position influence infrared atlas of each component position and the ambient temperature in the original infrared time-series image sequence, the original infrared time-series image sequence is corrected to obtain the chip module's own heat generation infrared time-series image sequence; S5: Supervise the heat generation of components in the self-heating infrared time-series image sequence based on the time-series model.

2. The method for automatically controlling and supervising chip functions according to claim 1, characterized in that: The step S2 includes: numbering N key components in the chip module and obtaining corresponding position information ; Set M discrete local target heating temperature values and the overall temperature, wherein the overall temperature is the operating environment temperature of the chip module; a double loop structure is used to traverse N key component positions and M discrete local target heating temperature values to obtain M temperature-position influence infrared images of each component, forming a set of temperature-position influence infrared images of all components.

3. The method for automatically controlling and supervising chip functions according to claim 1, wherein: The temperature-position effect infrared image collection of all components using a double loop structure specifically includes: Outer loop: Select the component position to be calibrated , , the central control system is based on Activate the local heating unit at its corresponding position; Inner loop: For the currently activated local heating unit, select the target heating temperature value in turn ; Adjust the power applied to the local heating unit so that the temperature of the corresponding area gradually increases and stabilizes at the target heating temperature value.

4. The method for automatically controlling and supervising chip functions according to claim 3, wherein: In the inner cycle, under thermal steady-state conditions, the temperature distribution of the entire chip module is collected when each component is heated to the target temperature value; Once the thermal steady state is confirmed, the central control system triggers the infrared thermal imager to capture and store an infrared image. ; Records the time when the kth element is heated to The temperature distribution of the entire chip module is .

5. The method for automatically controlling and supervising chip functions according to claim 4, wherein: The inner cycle also includes cooling and starting the re-circulation process: completing an infrared image After the data is collected, the central control system turns off the current local heating unit. When the entire chip module cools down to the operating environment temperature, the inner cycle continues and the same component position Apply the next target ramp temperature value , repeat heating, steady-state judgment, infrared image acquisition and storage until all M target heating temperature values are completed; The outer loop also includes: when the element position After all the infrared images of the M target temperature values are collected, the outer loop enters the next iteration and selects the new component position , and repeat all the operations of the inner loop.

6. The method for automatically controlling and supervising chip functions according to claim 5, wherein: Start the detection and obtain the original infrared time-series image sequence of the chip module including; The global temperature control system of the incubator is used to adjust the overall ambient temperature inside the incubator to the operating ambient temperature. The test system applies power to the chip module and executes the predetermined workload, triggering the infrared thermal imager to continuously capture infrared thermal images of the chip module surface. The infrared thermal imager generates raw infrared image frames at the set frame rate. , i is the image frame number from the start time.

7. The method for automatically controlling and supervising chip functions according to claim 6, wherein: A triple loop structure is used to obtain the chip module's own heat generation infrared time-series image sequence. The first loop iteration includes extracting any original infrared image frame in the original infrared time-series image sequence. , generate a corresponding corrected image .

8. The method for automatically controlling and supervising chip functions according to claim 7, wherein: For each raw infrared image frame The correction process sets a second loop structure, which traverses all key elements to be corrected ; For the currently selected key element , get the key component in Corresponding position in The average observed temperature ; For each key component to be corrected The correction process sets a third loop structure that traverses all other key elements , , .

9. The method for automatically controlling and supervising chip functions according to claim 8, wherein: Correct each key element to be corrected based on the third loop structure include: For other key components , get the key component exist Corresponding position in The average observed temperature , access the temperature-position effect infrared map corresponding to the key components M infrared images, find Corresponding to the position in M infrared images The average temperature of the current frame The average observed temperature The closest one , calculated At temperature Time The resulting temperature rise is: ; in, represents the temperature rise effect of key component j on key component k in the i-th frame, Indicates the target temperature rise value of key component j The average temperature of the key component k position in the obtained temperature-position-affecting infrared image; Operating ambient temperature; The calculated key components For key components Thermal influence Added to the total external heat impact: ; in, is the key element of frame i The total temperature rise affected.

10. A system for automatically controlling and supervising chip functions, characterized in that: The system includes: Temperature box: Place the chip module in the temperature box and connect the test circuits; Local temperature rise control module: Based on the position information of each component in the chip module, it instructs the temperature chamber to perform local temperature rise, and obtains the infrared image set of the temperature-position influence of all components; Infrared time-series image sequence acquisition module: Start detection and use the infrared thermal imager to obtain the original infrared time-series image sequence of the chip module; Image correction module: Based on the temperature-position influence infrared atlas of each component position and the ambient temperature in the original infrared time-series image sequence, the original infrared time-series image sequence is corrected to obtain the chip module's own heat generation infrared time-series image sequence; Supervision module: Based on the time series model, the heat generation of the components in the self-heating infrared time series image sequence is supervised.

Images