Circuit performance evaluation method in extreme environment
By establishing the correspondence between component model parameters and extreme environmental factors and using the Monte Carlo method, the problems of probabilistic changes in circuit parameters and multi-factor coupling under extreme environments were solved, realizing probabilistic evaluation of circuit performance and reliability design.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- HARBIN INST OF TECH
- Filing Date
- 2025-12-31
- Publication Date
- 2026-04-10
AI Technical Summary
Existing technologies are unable to effectively simulate the probabilistic changes in circuit parameters and the coupling effects of multiple factors under extreme environments, resulting in complex and unpredictable circuit behavior and a lack of probabilistic characterization in the evaluation results.
Establish the correspondence between component model parameters and extreme environmental factors, combine the probability distribution of multiple influencing factors and Monte Carlo method to perform random sampling, generate simulation input files and execute circuit simulation, and apply preset fault mode judgment rules to generate conditional probability tables.
It achieves nondeterministic evaluation of circuit performance and outputs probabilistic data, providing direct and quantitative decision-making basis for the reliability design of electronic systems in extreme environments, and solving the shortcomings of traditional methods in evaluation under complex environments and multivariate coupling.
Smart Images
Figure CN121835566A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of extreme environment circuit analysis technology, and more specifically, to a method for evaluating circuit performance under extreme environments. Background Technology
[0002] In extreme environments such as radiation and extreme temperatures, the parameters of components in electronic circuits undergo significant and unpredictable changes. These changes often exhibit a probability distribution rather than fixed values. Furthermore, the signal and power coupling effects between modules further increase the complexity of system behavior, making the performance degradation and failure modes of circuits under extreme environments highly uncertain and multi-factor-driven, thus rendering circuit behavior highly complex and difficult to predict.
[0003] In related technologies, currently widely used circuit simulation methods, such as SPICE simulation based on fixed-parameter models, are essentially deterministic analyses. Their simulation results depend on idealized component models and preset environmental conditions. Alternatively, parameter scanning may be used, but this is mostly limited to discretization analysis of single variables, making it difficult to effectively simulate circuit behavior under the random coupling of multiple environmental variables, and to characterize the probabilistic changes in component parameters themselves. Furthermore, existing simulation tools typically only output electrical signal waveforms or data curves, making it difficult to directly combine them with complex fault criteria accumulated from engineering experience for automated judgment and statistics. This results in evaluation results lacking the probabilistic representation of actual reliability. Summary of the Invention
[0004] The problem addressed by this invention is how to improve the ability to characterize probabilistic parameter changes and multi-factor coupling effects in circuit performance evaluation under extreme environments.
[0005] To address the above problems, this invention provides a method for evaluating circuit performance under extreme environments, comprising: Establish the correspondence between at least one model parameter of a selected component in the target circuit and at least one extreme environmental factor; Multiple influencing factors affecting the performance of the target circuit are identified, and a probability distribution of the values of each influencing factor is established, wherein the influencing factors include at least one of the extreme environmental factors; Based on the Monte Carlo method, multiple random samplings are performed according to the probability distribution, and each random sampling yields a set of specific values for the influencing factors. The values of the extreme environmental factors are obtained by filtering based on the specific values of each group of influencing factors, and the specific values of the model parameters are determined by the correspondence of the extreme environmental factors. Based on the specific values of the model parameters, a simulation input file corresponding to the target circuit is generated, and circuit simulation is performed based on the simulation input file to obtain simulation results; The simulation results are judged according to the preset fault mode judgment rules to determine the correspondence between the specific value corresponding to the current sampling and the judgment result. Based on all the aforementioned correspondences, a conditional probability table is generated that reflects the probability of circuit failure under different combinations of influencing factor values.
[0006] Optionally, establishing the correspondence between at least one model parameter of a selected component in the target circuit and at least one extreme environmental factor includes: Obtain experimental data of the selected component under the influence of at least one extreme environmental factor; Based on the experimental data, an interpolation process is used to establish the correspondence between the model parameters and the values of the extreme environmental factors.
[0007] Optionally, the step of determining multiple influencing factors affecting the performance of the target circuit and establishing a probability distribution for the values of each influencing factor, wherein the influencing factors include at least one of the extreme environmental factors, including: The influencing factors include at least two of the following: total ionized absorbed dose, temperature, power supply voltage, and input level. Define one or more states for each of the influencing factors, and establish a probability distribution for each state, wherein establishing the probability distribution includes establishing a uniform distribution, a normal distribution, or an empirical distribution based on historical data statistics.
[0008] Optionally, the Monte Carlo method involves multiple random samplings based on the probability distribution, with each random sampling yielding a set of specific values for the influencing factors, including: The Monte Carlo method is used to conduct multiple independent experiments. Each experiment includes: randomly generating a value based on the probability distribution of each of the influencing factors, thereby obtaining a set of specific values corresponding to all the influencing factors; wherein the number of independent experiments is the same as the number of sets of specific values obtained.
[0009] Optionally, the step of filtering based on the specific values of each group of influencing factors to obtain the values of the extreme environmental factors, and determining the specific values of the model parameters through the correspondence of the extreme environmental factors, includes: From the specific values of the current set of influencing factors, based on the attribute type of the influencing factors, specific values of the type of extreme environmental factors are selected from the current set of specific values; The specific values of the extracted extreme environmental factors are input into the corresponding relationship, and the specific values of the model parameters are output.
[0010] Optionally, generating a simulation input file corresponding to the target circuit based on the specific values of the model parameters, and performing circuit simulation based on the simulation input file to obtain simulation results, includes: The specific values of the model parameters are written into a reference file used to describe the target circuit structure to generate the simulation input file; The simulation input file is read and executed using the SPICE simulator, and the simulation results containing circuit node voltages and / or branch current data are output.
[0011] Optionally, the step of judging the simulation results for faults according to preset fault mode judgment rules, and determining the correspondence between the specific value corresponding to the current sample and the judgment result, includes: Extract at least one key circuit performance data from the simulation results; The key circuit performance data is compared with a preset normal operating threshold range, and the circuit is determined to be in a normal or faulty state based on the comparison result to obtain the judgment result. Establish an association between the judgment result and a set of specific values corresponding to the current sampling to determine a correspondence.
[0012] Optionally, generating a conditional probability table reflecting the probability of circuit failure under different combinations of influencing factor values based on all the aforementioned correspondences includes: All the correspondences are classified according to the specific set of values contained therein to obtain multiple combinations of input conditions; For each type of input condition combination, the number of times the fault state occurs in all corresponding judgment results is counted, and the proportion of this number to the total number of judgment results in that type is calculated as the probability of circuit fault occurrence under that type of input condition combination. The conditional probability table is generated by summarizing various combinations of input conditions and their corresponding failure probabilities.
[0013] Optionally, the key circuit performance data is the output voltage of at least one node in the target circuit.
[0014] Optionally, the selected component includes a bipolar junction transistor (BJT), and the model parameters include the current amplification factor of the BJT.
[0015] The circuit performance evaluation method under extreme environments of this invention establishes a correspondence between component model parameters and extreme environmental factors. By combining this with probability distributions established for multiple influencing factors and Monte Carlo random sampling, circuit simulation can dynamically reflect the probabilistic changes of component parameters under environmental stress and cover complex scenarios involving the random coupling of multiple factors. This overcomes the inherent limitations of traditional methods, which, due to their use of fixed parameters and univariate scanning, cannot characterize parameter uncertainties and the interactions of multiple factors.
[0016] Furthermore, by performing automated circuit simulation based on dynamically generated model parameters and applying preset fault judgment rules to batch-judge and correlate the simulation results, a conditional probability table is ultimately generated that quantifies the probability of fault occurrence under different environmental and operating condition combinations. This transforms the evaluation results from traditional deterministic waveforms or curves into probabilistic data that directly characterizes circuit reliability, solving the fundamental problem of the lack of probabilistic characterization capabilities in existing technologies.
[0017] In summary, this invention transforms traditional deterministic circuit simulation into a nondeterministic probabilistic evaluation. Furthermore, the final output conditional probability table provides a direct and quantitative basis for the reliability design, hardening verification, and lifetime prediction of electronic systems in extreme environments. It effectively realizes the nondeterministic evaluation of circuit performance under extreme conditions and makes up for the shortcomings of traditional deterministic analysis in dealing with complex environments and multivariate coupling. Attached Figure Description
[0018] Figure 1 This is a flowchart illustrating the circuit performance evaluation method under extreme conditions according to an embodiment of the present invention. Figure 2 This is the experimental data table for embodiment 2N2222 of the present invention; Figure 3 This is an interpolation curve of the Bf parameter as a function of total ionization absorbed dose in an embodiment of the present invention; Figure 4 This is an experimental circuit diagram of an embodiment of the present invention; Figure 5 This is a table showing the status and distribution of various influencing factors in this embodiment of the invention; Figure 6 This is a failure mode discrimination script for an embodiment of the present invention; Figure 7 This is the final conditional probability table for this embodiment of the invention. Detailed Implementation
[0019] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Although some embodiments of the present invention are shown in the drawings, it should be understood that the present invention can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of the present invention. It should be understood that the accompanying drawings and embodiments of the present invention are for illustrative purposes only and are not intended to limit the scope of protection of the present invention.
[0020] It should be understood that the various steps described in the method embodiments of the present invention may be performed in different orders and / or in parallel. Furthermore, the method embodiments may include additional steps and / or omit the steps shown. The scope of the present invention is not limited in this respect.
[0021] The term "comprising" and its variations as used herein are open-ended, meaning "including but not limited to"; the term "based on" means "at least partially based on"; the term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments"; and the term "optionally" means "optional embodiments". Definitions of other terms will be given in the following description. It should be noted that the concepts of "first," "second," etc., mentioned in this invention are used only to distinguish different devices, modules, or units, and are not intended to limit the order of functions performed by these devices, modules, or units or their interdependencies.
[0022] It should be noted that the terms "a" and "a plurality of" used in this invention are illustrative rather than restrictive. Those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".
[0023] It should be noted that the information (including but not limited to user device information, user personal information, etc.), data (including but not limited to data used for analysis, data stored, data displayed, etc.) and signals involved in this application are all authorized by the user or fully authorized by all parties. The collection, use and processing of related data must comply with the relevant laws, regulations and standards of the relevant countries and regions, and corresponding operation portals are provided for users to choose to authorize or refuse.
[0024] Combination Figure 1 As shown in the figure, the present invention provides a method for evaluating circuit performance under extreme environments, comprising: Establish the correspondence between at least one model parameter of a selected component in the target circuit and at least one extreme environmental factor.
[0025] Specifically, a correspondence is established between at least one model parameter of a selected component in the target circuit and at least one extreme environmental factor. This is achieved by obtaining experimental data on the performance degradation of the component under specific extreme environments, such as total ionizing radiation, and processing these discrete data points using interpolation methods to form a functional relationship or lookup table describing the continuous change of model parameters, such as the current amplification factor Bf of a bipolar transistor, with environmental factors, such as total ionizing dose TID.
[0026] For example, in this embodiment, extreme environments such as radiation environments are used, where the total ionizing absorbed dose (TID) represents the total ionizing radiation dose absorbed by the electronic device in the radiation field. For instance, in space applications, electronic devices are exposed to cosmic rays and high-energy particles, leading to parameter degradation in semiconductor devices; for example, the current amplification factor (Bf) of a bipolar transistor decreases with increasing TID. Or, single-event effects (SEE): in radiation environments, high-energy particles may collide with semiconductor devices, causing transient pulses or permanent damage, such as single-event upsets (SEUs) or single-event burn-outs (SEBs). Another example is temperature environments, such as extreme high and low temperatures. Temperature is one of the key factors affecting the performance of electronic devices. Extreme low temperatures (e.g., -55°C) may cause changes in material properties, such as a decrease in carrier mobility in semiconductors; extreme high temperatures (e.g., 125°C or higher) may cause thermal runaway or accelerated aging of devices. Finally, power supply voltage fluctuations may occur; in extreme environments, the power supply system may be disturbed, leading to voltage fluctuations and instability. For example, in aerospace or military applications, the power supply voltage may fluctuate between 11.5V and 12.5V, which can affect the normal operation of the circuit and even cause malfunctions.
[0027] Multiple influencing factors affecting the performance of the target circuit are identified, and a probability distribution of the values of each influencing factor is established, wherein the influencing factors include at least one of the extreme environmental factors.
[0028] Specifically, based on the actual operating conditions of the circuit and historical environmental data, key factors such as total ionization absorbed dose, operating temperature, power supply voltage and input signal level are selected. Different state levels or numerical ranges are divided for each factor, and each state is assigned an occurrence probability based on statistics or experience, such as uniform distribution or normal distribution, thereby constructing a multi-factor joint probability model.
[0029] Based on the Monte Carlo method, multiple random samplings are performed according to the probability distribution, and each random sampling yields a set of specific values for the influencing factors.
[0030] Specifically, multiple random samplings are performed based on the Monte Carlo method to obtain the specific values of multiple influencing factors. Through a computer program, a large number of combinations of values that conform to the statistical law are generated independently and randomly according to the probability distribution of the above factors. Each sampling simulates a possible combination of environmental and working conditions.
[0031] The values of the extreme environmental factors are obtained by filtering based on the specific values of each group of influencing factors, and the specific values of the model parameters are determined by the correspondence of the extreme environmental factors.
[0032] Specifically, from each complete parameter combination of sampling, the values of extreme environmental factors, such as total ionizing dose (TID), are identified and input into the correspondence established in the first step. The corresponding accurate model parameter values under this environmental condition are then queried or calculated, such as the specific Bf value (Beta Forward, forward current amplification factor).
[0033] Based on the specific values of the model parameters, a simulation input file corresponding to the target circuit is generated, and circuit simulation is performed based on the simulation input file to obtain simulation results.
[0034] Specifically, the calculated model parameter values are updated or replaced in the standard SPICE netlist description file of the target circuit to form a specific simulation input file corresponding to this sampling. Then, the SPICE simulation engine is called to perform calculations, and the voltage, current, and other time-domain or frequency-domain response waveform data of the output circuit are used as simulation results.
[0035] The simulation results are used to determine faults according to the preset fault mode judgment rules, and the correspondence between the specific value corresponding to the current sampling and the judgment result is determined.
[0036] Specifically, based on the circuit design specifications, such as the normal output voltage range, an automated script is written to analyze the simulation result waveform, determine whether it exceeds the tolerance, and thus obtain a judgment conclusion of normal or fault. This judgment result is then associated with and stored in conjunction with the sampling value combination on which the simulation is based.
[0037] Based on all the aforementioned correspondences, a conditional probability table is generated that reflects the probability of circuit failure under different combinations of influencing factor values.
[0038] Specifically, a large number of simulation records are statistically analyzed, and records with the same or similar combinations of influencing factors are categorized. The proportion of times the circuit is judged as faulty under each combination is calculated to the total number of times. Finally, a table showing the probability of circuit fault occurrence under different combinations of environment and operating conditions is compiled to complete the probabilistic evaluation.
[0039] This embodiment of the circuit performance evaluation method under extreme environments establishes a correspondence between component model parameters and extreme environmental factors. By combining this with probability distributions established for multiple influencing factors and Monte Carlo random sampling, the circuit simulation can dynamically reflect the probabilistic changes of component parameters under environmental stress and cover complex scenarios involving the random coupling of multiple factors. This overcomes the inherent limitations of traditional methods, which, due to their use of fixed parameters and univariate scanning, cannot characterize parameter uncertainties and the interactions of multiple factors.
[0040] Furthermore, by performing automated circuit simulation based on dynamically generated model parameters and applying preset fault judgment rules to batch-judge and correlate the simulation results, a conditional probability table is ultimately generated that quantifies the probability of fault occurrence under different environmental and operating condition combinations. This transforms the evaluation results from traditional deterministic waveforms or curves into probabilistic data that directly characterizes circuit reliability, solving the fundamental problem of the lack of probabilistic characterization capabilities in existing technologies.
[0041] In summary, this embodiment transforms traditional deterministic circuit simulation into a nondeterministic probabilistic evaluation. Furthermore, the final output conditional probability table provides a direct and quantitative basis for the reliability design, hardening verification, and lifetime prediction of electronic systems in extreme environments. It effectively realizes the nondeterministic evaluation of circuit performance under extreme conditions and makes up for the shortcomings of traditional deterministic analysis in dealing with complex environments and multivariate coupling.
[0042] Optionally, establishing the correspondence between at least one model parameter of a selected component in the target circuit and at least one extreme environmental factor includes: Obtain experimental data of the selected component under the influence of at least one extreme environmental factor; Based on the experimental data, an interpolation process is used to establish the correspondence between the model parameters and the values of the extreme environmental factors.
[0043] Specifically, by reviewing publicly available literature on device radiation or high / low temperature testing, or by designing and executing controlled environmental simulation experiments, key electrical characteristic parameters of the device, such as the degradation data of the current amplification factor Bf of a bipolar transistor, are measured and recorded at multiple discrete extreme environmental factor levels of varying intensities, such as different total ionizing dose points. This data is then compiled into a discrete dataset table characterizing the correspondence between environmental factor values and model parameter values. Based on the experimental data, numerical analysis methods, such as linear interpolation, spline interpolation, or least squares fitting, are used to fit the discrete data points into a continuous or piecewise continuous curve or function. This allows for the prediction or calculation of accurate model parameter values based on any extreme environmental factor input value within the experimental data range, and the functional relationship is encapsulated into a parameterized model that can be called by simulation scripts.
[0044] In this optional embodiment, by using experimental data and interpolation methods to construct a quantitative correspondence between model parameters and environmental factors, the constant component parameters in traditional simulation are transformed into variables that dynamically adjust with environmental input values. This provides a dynamically callable parameter model foundation that reflects the real physical degradation laws for subsequent Monte Carlo sampling and simulation. It solves the key defect in traditional deterministic simulation that the component model cannot characterize the probabilistic drift of parameters under extreme environments, and realizes the transformation of the simulation model from static idealization to dynamic environmental adaptability.
[0045] Optionally, the step of determining multiple influencing factors affecting the performance of the target circuit and establishing a probability distribution for the values of each influencing factor, wherein the influencing factors include at least one of the extreme environmental factors, including: The influencing factors include at least two of the following: total ionized absorbed dose, temperature, power supply voltage, and input level. Define one or more states for each of the influencing factors, and establish a probability distribution for each state, wherein establishing the probability distribution includes establishing a uniform distribution, a normal distribution, or an empirical distribution based on historical data statistics.
[0046] Specifically, based on the actual application scenario and failure mechanism analysis of the target circuit, physical and electrical quantities that have a significant impact on circuit performance are selected. For example, for space application circuits, total ionization absorbed dose and temperature are key environmental factors, while power supply voltage fluctuations and input signal level tolerances are key electrical conditions that determine whether the circuit functions normally. Based on prior knowledge or measured data, the continuous value range of each factor is divided into engineering-significant intervals, such as dividing the total ionization absorbed dose into low, medium, and high intervals, or discretizing it into several typical level values. Furthermore, the most likely probability distribution type is specified for each state. For example, a uniform distribution is used for states lacking prior information, and a normal distribution is used for variables with a central tendency, such as temperature. Alternatively, the empirical probability distribution can be obtained directly from a large amount of historical observation data, thus completing a comprehensive mathematical model of the joint characteristics of multi-factor uncertainty.
[0047] In this optional embodiment, by dividing multiple key influencing factors, such as total ionization dose, temperature, and power supply voltage, into different states and establishing applicable probability distributions for them, the parameters that are usually regarded as deterministic values or single boundary conditions in traditional analysis are systematically transformed into a series of random variables that conform to real statistical laws. This provides a joint probability model that can accurately reflect the uncertainty of each factor and the complex coupling relationship between them for subsequent Monte Carlo sampling. It lays the data foundation for generating a large number of statistically significant combinations of environments and operating conditions covering real-world scenarios, and solves the problem that traditional deterministic simulation or univariate scanning cannot characterize the random interaction of multiple factors.
[0048] Optionally, the Monte Carlo method involves multiple random samplings based on the probability distribution, with each random sampling yielding a set of specific values for the influencing factors, including: The Monte Carlo method is used to conduct multiple independent experiments. Each experiment includes: randomly generating a value based on the probability distribution of each of the influencing factors, thereby obtaining a set of specific values corresponding to all the influencing factors; wherein the number of independent experiments is the same as the number of sets of specific values obtained.
[0049] Specifically, the computer program executes cyclically under a preset total number of experiments. In each independent experimental cycle, for each influencing factor with an established probability distribution, its corresponding random number generation algorithm is invoked. For example, the inverse transformation sampling method based on uniform distribution random number transformation independently extracts a specific value from its specific probability distribution, and combines the values extracted for all influencing factors in this cycle to form a set of specific values representing the unique combination of conditions for this experiment. Simultaneously, since each complete independent experiment logically produces only one set of specific values corresponding to all influencing factors, the total number of experiments executed by the program directly determines the total number of complete parameter combinations generated that can be used for subsequent simulations. This explicit correspondence in this embodiment ensures the controllability of the sampling scale and the completeness of the results.
[0050] It should be noted that the Monte Carlo method is applied in this embodiment. The total number of simulations is preset by the computer program, for example, one million times. Then, in each independent simulation experiment, based on the established probability distribution of each influencing factor, such as the uniform distribution of total ionizing absorbed dose in the corresponding state interval and the normal distribution of temperature, a specific value is extracted for each factor by calling a random number generator. For example, a single sampling yields TID=12.5 krad, temperature=85℃, power supply voltage=11.8V, and input level=0.2V. These values are combined to form an input parameter set representing an extreme combination of environmental and operating conditions. By repeating this process up to a predetermined number of times, various random and coupled nondeterministic operating conditions that the circuit may face in practice can be simulated on a large scale.
[0051] In this optional embodiment, by applying the Monte Carlo method for large-scale independent sampling, the static probability distribution of each influencing factor established in the previous steps can be efficiently and completely transformed into massive amounts of specific scenario data that can be directly used for subsequent simulations. The design that the number of samplings is strictly consistent with the number of parameter groups obtained ensures that each independent experiment corresponds to a unique and clear combination of environment and operating conditions, thereby guaranteeing the completeness of the sampling process and the traceability of the results in terms of algorithm logic, and laying a rigorous data foundation for subsequent accurate circuit simulation and reliable statistical evaluation.
[0052] Optionally, the step of filtering based on the specific values of each group of influencing factors to obtain the values of the extreme environmental factors, and determining the specific values of the model parameters through the correspondence of the extreme environmental factors, includes: From the specific values of the current set of influencing factors, based on the attribute type of the influencing factors, specific values of the type of extreme environmental factors are selected from the current set of specific values; The specific values of the extracted extreme environmental factors are input into the corresponding relationship, and the specific values of the model parameters are output.
[0053] Specifically, the program identifies the category label assigned to each influencing factor during its initial definition. After obtaining a complete parameter set through each sampling, it automatically iterates through and compares the labels of each parameter, extracting only the parameters labeled as extreme environmental factors and their values, which are then used as input for subsequent parameter calculations. The extracted values of one or more environmental factors are substituted into the correspondence model established in the first step and encapsulated as functions or lookup tables. A calculation or query operation is then performed, resulting in a definite model parameter value that matches the specific environmental conditions, thus completing the transformation from random environmental input to defined simulation model parameters.
[0054] In this optional embodiment, an automated data interface is achieved between Monte Carlo random sampling results and SPICE simulation model parameters through two consecutive steps: attribute filtering and relationship mapping. This system can automatically identify and extract key environmental variables from mixed parameter sets and convert them into precise component parameters required for simulation based on predefined physical relationships. This ensures that each subsequent deterministic SPICE simulation is based on the currently randomly generated environmental conditions, guaranteeing the accurate operation of the entire probabilistic evaluation process.
[0055] Optionally, generating a simulation input file corresponding to the target circuit based on the specific values of the model parameters, and performing circuit simulation based on the simulation input file to obtain simulation results, includes: The specific values of the model parameters are written into a reference file used to describe the target circuit structure to generate the simulation input file; The simulation input file is read and executed using the SPICE simulator, and the simulation results containing circuit node voltages and / or branch current data are output.
[0056] Specifically, a script parses a baseline file, typically a standard SPICE netlist file containing circuit topology, component models, and initial parameters. It locates the parameter row corresponding to the selected component model and replaces existing parameter values with currently calculated model parameter values, such as Bf=150, or inserts them as new parameters. This generates a new, updated SPICE netlist file that can be directly read by the simulator. The script then automatically calls a SPICE simulator, such as an HSPICE simulator, submitting the newly generated netlist file as input for calculation. The simulator solves the circuit equations based on the circuit description and component parameters in the netlist, performing transient, DC, or AC analysis. It then outputs the calculated node voltage waveforms, branch current data, or other specified output variables in standard data file formats, such as .tr0 or .csv. These data files constitute the simulation results.
[0057] Optionally, the step of judging the simulation results for faults according to preset fault mode judgment rules, and determining the correspondence between the specific value corresponding to the current sample and the judgment result, includes: Extract at least one key circuit performance data from the simulation results; The key circuit performance data is compared with a preset normal operating threshold range, and the circuit is determined to be in a normal or faulty state based on the comparison result to obtain the judgment result. Establish an association between the judgment result and a set of specific values corresponding to the current sampling to determine a correspondence.
[0058] Specifically, a script is written to automatically parse the standard waveform data file output by the SPICE simulator. Based on predefined circuit observation nodes or device pin names, it locates and reads the voltage or current values at specific simulation moments, such as steady-state values or values meeting specific conditions, such as maximum or minimum values. The script presets upper and lower limits for numerical ranges representing normal circuit function. By logically comparing the extracted performance data with these thresholds (e.g., determining if the data is within the range), and automatically assigning a discrete state identifier, such as NORMAL or BROKEN, based on the comparison result, the conversion from continuous simulation data to binary fault states is completed.
[0059] Furthermore, a structured data record is created in the program memory or an external database, and this record contains at least two core fields: one field stores the specific values of a complete set of influencing factors that triggered the simulation and the current sampling, and the other field stores the fault judgment result obtained from the simulation. The program logically binds and persistently stores these two fields to form a complete input-output mapping record.
[0060] In this optional embodiment, by automatically extracting key performance data from simulation waveforms and objectively comparing it with preset thresholds, the experience-based, qualitative fault judgment is transformed into an automated, repeatable, quantitative judgment process. By structurally associating and recording the result of each judgment with the precise input conditions that produced that result, standardized input-output sample data required for subsequent statistical calculations is generated. This provides an accurate, complete, and traceable data foundation for the final generation of the conditional probability table, solving the problems of difficulty in objectively incorporating fault modes and the inability to statistically analyze evaluation results in traditional methods.
[0061] Optionally, generating a conditional probability table reflecting the probability of circuit failure under different combinations of influencing factor values based on all the aforementioned correspondences includes: All the correspondences are classified according to the specific set of values contained therein to obtain multiple combinations of input conditions; For each type of input condition combination, the number of times the fault state occurs in all corresponding judgment results is counted, and the proportion of this number to the total number of judgment results in that type is calculated as the probability of circuit fault occurrence under that type of input condition combination. The conditional probability table is generated by summarizing various combinations of input conditions and their corresponding failure probabilities.
[0062] Specifically, the program script traverses all stored input-output records, grouping records with identical values of influencing factors into the same set. Alternatively, for statistical purposes, the specific values of each influencing factor are discretized according to a preset precision, such as rounding or the state range they belong to, such as the temperature belonging to the high-temperature range. Records with the same combination of processed values are then categorized, thus forming multiple data subsets identified by different combinations of input conditions.
[0063] Based on the classification, the program reads all judgment result fields within each subset, counts the number of records marked as fault states, and then divides this count by the total number of records in that subset to calculate the frequency of the fault state. This frequency is defined as the conditional probability of a circuit fault occurring under a specific combination of input conditions. The program then represents each calculated combination of input conditions, for example, as a discretized numerical vector or state label vector, along with its corresponding conditional probability value. This information is organized and written into a structured two-dimensional table or matrix. The rows or columns of this table clearly display different combinations of input conditions and are associated with their corresponding unique fault probability values. Finally, a complete conditional probability table document is generated for viewing or exporting.
[0064] In this optional embodiment, by automatically classifying, statistically analyzing, and calculating a large number of input-output records, the large-scale discrete simulation data obtained in the aforementioned steps is systematically refined, compressed, and organized into a structured conditional probability table. This table clearly reveals the statistical laws of circuit failure under different extreme environmental and operating condition combinations with quantified probability values, thereby transforming the simulation results into deterministic probabilistic data that can be directly used by reliability engineers to support design decisions and risk assessments, thus solving the problem that existing assessment results lack probabilistic characterization capabilities.
[0065] Optionally, the key circuit performance data is the output voltage of at least one node in the target circuit.
[0066] In this optional embodiment, when writing the fault diagnosis script in the early stages, the signals read and analyzed from the SPICE simulation result file are explicitly specified as specific circuit nodes, such as the voltage value at the output of the inverter at a specific simulation moment, such as the voltage value at the end of the transient simulation. This value directly characterizes whether the core logic function or driving capability of the circuit is normal. By explicitly defining the key performance data as the node output voltage, the original waveform data generated by the simulation is directly linked to the most intuitive and critical circuit function indicators, namely the correct logic level or voltage amplitude. This makes the automated fault diagnosis rules have a clear, objective, and easily verifiable engineering basis, thereby ensuring a strong correlation between the probability assessment results and the actual failure mode of the circuit.
[0067] Optionally, the selected component includes a bipolar junction transistor (BJT), and the model parameters include the current amplification factor of the BJT.
[0068] In this optional embodiment, based on the topology of the target circuit and its sensitivity to the radiation environment, a specific type of transistor is selected as the key degradation device during the modeling stage. A quantitative degradation relationship is established between the forward current amplification factor (Bf), the core parameter that directly determines the amplification capability in its SPICE model, such as the Gummel-Poon model, and extreme environmental factors such as the total ionization dose.
[0069] In a preferred embodiment of the present invention, a typical BJT device, 2N2222, was selected. Data was collected from publicly available literature on its total ionization dose experiments, and the data is presented in the table below. Figure 2 As shown, interpolation was performed using a common and readily available method to obtain the curve of the Bf parameter of the BJT element as a function of the total ionizing dose. The interpolation curve is shown in Figure 1. Figure 3 As shown.
[0070] Build simulation circuits, such as Figure 4 As shown, the core circuit carrier used for simulation verification in this embodiment of the invention is the basic hardware model for subsequent multi-factor coupling analysis, fault mode identification, and parameter sampling and replacement. Combined with... Figure 4 As shown, the circuit type is a 12V powered CMOS level inverter. Its core function is to reverse the input and output levels, such as a high input level corresponding to a low output level, and vice versa. Specifically, it includes a BJT crystal (2N2222), which is an NPN bipolar junction transistor (BJT). As the core switch or amplification unit in the circuit, its key parameter Bf and current amplification factor decrease with changes in the total ionization absorbed dose TID. Figure 3As shown in the interpolation curve, the radiation effect directly affects the switching characteristics and output voltage stability of the inverter, and is the main target of radiation effects in extreme environments. It also includes CMOS field-effect transistors (MOSFETs), consisting of one PMOS (P-channel) and one NMOS (N-channel), forming the core structure of the CMOS inverter. The PMOS is connected to the +12V power supply, and the NMOS is connected to ground; the gates of both transistors are connected together as the input level terminal, and the drains are connected together as the output level terminal. During operation, the high or low input level controls the conduction / cut-off of the PMOS / NMOS, realizing the inversion function of high input level - low output level and low input level - high output level, which is the core functional carrier of the 12V CMOS inverter. A DC power supply (+12V) serves as a DC regulated power supply, providing a stable 12V operating voltage for the entire CMOS inverter, and is the energy source for normal circuit operation. The input signal source is a DC level signal source or a pulse level signal source, used to provide the input level (high / low level) of the inverter, and is used to test whether the output response of the circuit is normal under different input states. Load components, such as resistors R1 or R2, are typically current-limiting or load resistors, such as fixed resistors connected in series or parallel at the output. They are used to protect CMOS transistors or BJT transistors from short circuits at the output, preventing damage to the devices.
[0071] Based on existing actual circuit operating conditions and environmental effects, key influencing factors were selected. Here, four key factors were chosen: Total Ionized Dose (TID), thermostat temperature, 12V power supply voltage, and input level. Different states for each factor were defined, and for each factor and state, a probability distribution was obtained based on experience or statistical analysis of actual data. The probability state table in this embodiment is as follows: Figure 5 As shown. The key variables affecting circuit performance under extreme environments, their specific states, and the probability distribution of each state are clearly defined. The selected core influencing factors include three main categories: environment, power supply, and input signal, reflecting actual operating conditions in extreme environments. Total ionizing absorbed dose (TID), a core indicator of radiation effects, directly affects the Bf parameter of 2N2222; the thermostat temperature is a simulated variable for extreme high and low temperature environments; the 12V power supply voltage represents the fluctuation variable of the power supply system, and its stability is easily affected under extreme environments; the input level represents the state of the circuit's input signal and affects the output response of the level inverter. For example... Figure 5As shown, each factor is categorized into states, such as temperature, which may be divided into extreme / normal temperature states like -55℃, 25℃, and 125℃. The probability distribution for each state is derived based on experience or statistical analysis of actual data, such as a 20% probability of high-temperature states and a 15% probability of high-dose radiation states. The program script will randomly sample based on the probability distribution in this table. For example, if a sample selects the combination of TID=100krad, temperature=-55℃, power supply voltage=11.5V, and high-level input, it will provide specific operating parameters for a single simulation. Based on practical experience, operating conditions, actual needs, and specifications, the fault modes described in the scripting language are as follows: Figure 6 As shown, the fault determination rules are quantified and made executable using a scripting language. The circuit output voltage range is used as the fault threshold to clarify the boundary between normal and faulty states. At the same time, fault conditions are set. When the circuit output voltage exceeds the range of 8~12V, it is judged as abnormal. This abnormality is not only a performance failure of the circuit itself, but may also lead to the failure of subsequent related modules, such as the transmission relationship between abnormal output and module-level fault. After each sampling simulation, the script automatically reads the output voltage data and compares it with the threshold (8~12V). If it exceeds the threshold, the script records the fault occurrence and associates it with the combination of influencing factors of this sampling, such as "TID=200krad + temperature=125℃ results in output voltage=7.2V, which is judged as a fault." Automatic judgment by the script avoids the subjectivity of manual judgment and ensures that the fault judgment criteria of all simulation results are consistent. This indicates that for the circuit in this embodiment, an output voltage range outside 8~12V is abnormal and may lead to the failure of related modules. Using the program script, the number of samplings is set, and according to the above probability distribution, the program automatically performs batch random sampling and single simulations. For the total ionization absorbed dose, it is necessary to determine the appropriate value based on the TID obtained from the sampling. Figure 3 The corresponding relationships shown are used to rewrite and replace the Bf parameter in the SPICE simulation model of 2N2222; and the script mentioned in S3 is used to determine whether a fault has occurred and its specific mode, recording the results. 3,000,000 samples are taken, the results are statistically analyzed, identical input and output states are merged, and conditional probabilities are calculated separately to form a conditional probability table as the final result; this table represents the fault mode probability corresponding to various state combinations under comprehensive consideration of multi-factor coupling, environmental effects, and uncertainties. The final result is as follows: Figure 7 As shown.
[0072] While the present invention has been disclosed above, its scope of protection is not limited thereto. Those skilled in the art can make various changes and modifications without departing from the spirit and scope of the present invention, and all such changes and modifications will fall within the scope of protection of the present invention.
Claims
1. A method for evaluating circuit performance under extreme environments, characterized in that, include: Establish the correspondence between at least one model parameter of a selected component in the target circuit and at least one extreme environmental factor; Multiple influencing factors affecting the performance of the target circuit are identified, and a probability distribution of the values of each influencing factor is established, wherein the influencing factors include at least one of the extreme environmental factors; Based on the Monte Carlo method, multiple random samplings are performed according to the probability distribution, and each random sampling yields a set of specific values for the influencing factors. The values of the extreme environmental factors are obtained by filtering based on the specific values of each group of influencing factors, and the specific values of the model parameters are determined by the correspondence of the extreme environmental factors. Based on the specific values of the model parameters, a simulation input file corresponding to the target circuit is generated, and circuit simulation is performed based on the simulation input file to obtain simulation results; The simulation results are judged according to the preset fault mode judgment rules to determine the correspondence between the specific value corresponding to the current sampling and the judgment result. Based on all the aforementioned correspondences, a conditional probability table is generated that reflects the probability of circuit failure under different combinations of influencing factor values.
2. The circuit performance evaluation method under extreme environments according to claim 1, characterized in that, The establishment of the correspondence between at least one model parameter of a selected component in the target circuit and at least one extreme environmental factor includes: Obtain experimental data of the selected component under the influence of at least one extreme environmental factor; Based on the experimental data, an interpolation process is used to establish the correspondence between the model parameters and the values of the extreme environmental factors.
3. The circuit performance evaluation method under extreme environments according to claim 1, characterized in that, The process involves identifying multiple influencing factors affecting the performance of the target circuit and establishing a probability distribution for the values of each influencing factor. These influencing factors include at least one of the extreme environmental factors, including: The influencing factors include at least two of the following: total ionized absorbed dose, temperature, power supply voltage, and input level. Define one or more states for each of the influencing factors, and establish a probability distribution for each state, wherein establishing the probability distribution includes establishing a uniform distribution, a normal distribution, or an empirical distribution based on historical data statistics.
4. The circuit performance evaluation method under extreme environments according to claim 1, characterized in that, The Monte Carlo method involves multiple random samplings based on the probability distribution. Each random sampling yields a set of specific values for the influencing factors, including: The Monte Carlo method is used to conduct multiple independent experiments. Each experiment includes: randomly generating a value based on the probability distribution of each of the influencing factors, thereby obtaining a set of specific values corresponding to all the influencing factors; wherein the number of independent experiments is the same as the number of sets of specific values obtained.
5. The circuit performance evaluation method under extreme environments according to claim 1, characterized in that, The process of filtering based on the specific values of each group of influencing factors to obtain the values of the extreme environmental factors, and determining the specific values of the model parameters through the correspondence of the extreme environmental factors, includes: From the specific values of the current set of influencing factors, based on the attribute type of the influencing factors, specific values of the type of extreme environmental factors are selected from the current set of specific values; The specific values of the extracted extreme environmental factors are input into the corresponding relationship, and the specific values of the model parameters are output.
6. The circuit performance evaluation method under extreme environments according to claim 1, characterized in that, The step of generating a simulation input file corresponding to the target circuit based on the specific values of the model parameters, and performing circuit simulation based on the simulation input file to obtain simulation results, includes: The specific values of the model parameters are written into a reference file used to describe the target circuit structure to generate the simulation input file; The simulation input file is read and executed using the SPICE simulator, and the simulation results containing circuit node voltages and / or branch current data are output.
7. The circuit performance evaluation method under extreme environments according to claim 1, characterized in that, The step of judging the simulation results for faults according to preset fault mode judgment rules, and determining the correspondence between the specific value corresponding to the current sample and the judgment result, includes: Extract at least one key circuit performance data from the simulation results; The key circuit performance data is compared with a preset normal operating threshold range, and the circuit is determined to be in a normal or faulty state based on the comparison result to obtain the judgment result. Establish an association between the judgment result and a set of specific values corresponding to the current sampling to determine a correspondence.
8. The circuit performance evaluation method under extreme environments according to claim 1, characterized in that, The step of generating a conditional probability table reflecting the probability of circuit failure under different combinations of influencing factor values based on all the aforementioned correspondences includes: All the correspondences are classified according to the specific set of values contained therein to obtain multiple combinations of input conditions; For each type of input condition combination, the number of times the fault state occurs in all corresponding judgment results is counted, and the proportion of this number to the total number of judgment results in that type is calculated as the probability of circuit fault occurrence under that type of input condition combination. The conditional probability table is generated by summarizing various combinations of input conditions and their corresponding failure probabilities.
9. The circuit performance evaluation method under extreme environments according to claim 7, characterized in that, The key circuit performance data refers to the output voltage of at least one node in the target circuit.
10. The circuit performance evaluation method under extreme environments according to claim 1, characterized in that, The selected components include bipolar junction transistors (BJTs), and the model parameters include the current amplification factor of the BJTs.