Analog circuit parameter determination method and device, medium and product

Abstract

The invention discloses a parameter determination method and device of an analog circuit, a medium and a product, and relates to the technical field of integrated circuit design. Local performance feedback and global topology perception of performance index parameters are used as input parameters, and pertinence and generalization ability of parameter adjustment are remarkably improved. Through a strategy function, a parameter candidate set of a target analog circuit is obtained, through multi-group sampling processing, variance of strategy gradient estimation is reduced, and training stability and convergence speed are improved. Depending on a value network is not needed, and training resource overhead is reduced. And applying the parameter candidate set output by the strategy function to the simulation processing scene to realize the accuracy of simulation processing evaluation. The target performance indexes are extracted from the simulation result, optimization of the performance indexes is achieved, the multiple target performance indexes are processed according to the multi-target reward function, balance selection and rejection are achieved in the multiple performance indexes, the parameter combination corresponding to the highest advantage value is obtained, and the reliability of the analog circuit parameter determination process is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of integrated circuit design, and particularly to a method, device, medium and product for determining parameters of an analog circuit. Background Art

[0002] With the development of machine learning technology, the optimization method of analog circuit parameters based on machine learning is replacing the optimization based on human experience. In the optimization of analog circuit parameters by machine learning, most methods rely on a value network. The value network needs a large amount of sample data to estimate the value of the state or action so that the network convergence speed can be relatively accurate, which in turn leads to a large overhead of training resources and a slow convergence speed. At the same time, in the value network, the true value function is approximated iteratively, falling into a local optimal solution, and it is difficult to balance the performance indicators corresponding to multiple objectives in terms of the global optimal solution.

[0003] Therefore, how to save the training resource overhead, quickly and stably converge the speed, and improve local and global optimization is an urgent problem to be solved by those skilled in the art. Summary of the Invention

[0004] The purpose of the present invention is to provide a method, device, medium and product for determining parameters of an analog circuit to solve the technical problems of large training resource overhead, slow convergence speed, and difficulty in local and global optimization.

[0005] To solve the above technical problems, the present invention provides a method for determining parameters of an analog circuit, including: Obtaining the circuit source code and performance indicators corresponding to the target analog circuit; Determining the corresponding graph embedding vector based on the circuit source code, and inputting the graph embedding vector and the performance indicators into a policy function to obtain a parameter candidate set of the target analog circuit; Performing simulation processing on multiple parameter combinations of the parameter candidate set to obtain corresponding simulation results, and extracting corresponding target performance indicators according to the multiple simulation results; Processing the multiple target performance indicators according to a multi-objective reward function to obtain corresponding advantage values, and taking the parameter combination corresponding to the highest advantage value as the final parameter of the target analog circuit.

[0006] On the one hand, the determination process of the parameter candidate set includes: Pre-establishing a parameter database of the target analog circuit; wherein, the parameter database stores the parameters to be determined of the target analog circuit and their corresponding value ranges; Determining the value range of the output neuron output value of the policy function according to the parameter database; Perform multiple random samplings on the output value of the policy function according to the graph embedding vector and the performance metrics to generate the parameter candidate set.

[0007] On the other hand, perform simulation processing on multiple parameter combinations of the parameter candidate set to obtain corresponding simulation results, including: Apply multiple parameter combinations to the netlist in sequence; Call a simulator to perform circuit simulation processing on the netlist to obtain corresponding simulation results.

[0008] On the other hand, extract corresponding target performance metrics according to multiple simulation results, including: Determine the common performance metric corresponding to the frequency response according to the functional requirements, performance requirements, multiple parameter configurations, and sensitivity analysis of the target analog circuit among the performance metrics of the target analog circuit; Take the common performance metric as the target performance metric; Determine the corresponding target parameters according to the target performance metric among multiple simulation results.

[0009] On the other hand, process multiple target performance metrics according to a multi-objective reward function to obtain corresponding advantage values, including: Obtain the weight parameters corresponding to multiple target performance metrics; Substitute multiple target performance metrics, weight parameters, and target parameters into the multi-objective reward function to obtain the reward values corresponding to multiple parameter combinations; Determine the average value and variance value of the reward values of multiple parameter combinations; Take the average value and the variance value as the baseline, and perform advantage processing on multiple parameter combinations to obtain corresponding advantage values.

[0010] On the other hand, the target performance metrics are at least circuit gain metrics, bandwidth metrics, power consumption metrics, and linearity metrics; substitute multiple target performance metrics, weight parameters, and target parameters into the multi-objective reward function to obtain the reward values corresponding to multiple parameter combinations, including: Obtain the first weight parameter corresponding to the circuit gain metric, the second weight parameter corresponding to the bandwidth metric, the third weight parameter corresponding to the power consumption metric, and the fourth weight parameter corresponding to the linearity metric; Determine the first ratio value according to the circuit gain metric and the corresponding target parameters; Determine the second ratio value according to the bandwidth metric and the corresponding target parameters; Determine the third ratio value according to the power consumption metric and the corresponding target parameters; Determine the fourth ratio value according to the linearity metric and the corresponding target parameters; Process the first ratio value and the first weight parameter, the second ratio value and the second weight parameter, the third ratio value and the third weight parameter, and the fourth ratio value and the fourth weight parameter to obtain the reward values corresponding to multiple parameter combinations.

[0011] On the other hand, processing the first ratio value and the first weight parameter, the second ratio value and the second weight parameter, the third ratio value and the third weight parameter, and the fourth ratio value and the fourth weight parameter to obtain the reward values corresponding to multiple parameter combinations includes: Process the first ratio value and the first weight parameter to obtain a first reward value; Process the second ratio value and the second weight parameter to obtain a second reward value; Process the third ratio value and the third weight parameter to obtain a third reward value; Process the fourth ratio value and the fourth weight parameter to obtain a fourth reward value; Add the first reward value and the second reward value to obtain a fifth reward value; Subtract the third reward value and the fourth reward value from the fifth reward value in sequence to obtain the final reward value.

[0012] On the other hand, the sum of the first weight parameter and the second weight parameter is 1; the sum of the third weight parameter and the fourth weight parameter is 1.

[0013] On the other hand, the training process of the policy function includes: Call the initial policy function and input the current performance metric and the graph embedding vector into the initial policy function; where the current performance metric is the previous target performance metric obtained during the previous training of the policy function and is the initial performance metric during the first training; Output the current parameter candidate set corresponding to the initial policy function; Perform simulation processing on multiple parameter combinations of the current parameter candidate set to obtain the corresponding current simulation results, so as to extract the corresponding current target performance metrics; Process multiple current target performance metrics according to the multi-objective reward function to obtain the corresponding current advantage values; Fine-tune the initial policy function based on the current advantage values to obtain the policy function after parameter fine-tuning; When the current iteration number reaches the preset iteration number, then use the initial policy function as the final policy function; When the current iteration count has not reached the preset iteration count, the policy function after parameter fine-tuning is used as the new policy function and returned to the step of inputting the current performance metric and the graph embedding vector into the initial policy function until the current iteration count reaches the preset iteration count, and the final policy function is determined.

[0014] On the other hand, using the average value and the variance value as a baseline, performing an advantage process on multiple parameter combinations to obtain corresponding advantage values, including: Subtracting the current reward value from the average value to obtain a sixth reward value; Dividing the square root of the variance value by the sixth reward value to obtain the corresponding advantage value.

[0015] On the other hand, determining the corresponding graph embedding vector based on the circuit source code, including: Converting the circuit source code into a netlist; Constructing a circuit diagram of the target analog circuit based on the netlist; Invoking a graph neural network model, inputting the circuit diagram into the graph neural network model to obtain the corresponding graph embedding vector.

[0016] On the other hand, constructing a circuit diagram of the target analog circuit based on the netlist, including: Extracting the netlist nodes and netlist edge information corresponding to the netlist from the netlist, where the netlist nodes represent the components of the target analog circuit, and the netlist edge information represents the connection relationship between the components; Processing the netlist nodes and the netlist edge information through a graph data structure to obtain the circuit diagram.

[0017] To solve the above technical problems, the present invention also provides a device for determining parameters of an analog circuit, including: A memory for storing a computer program; A processor for implementing the steps of the method for determining parameters of an analog circuit as described above when executing the computer program.

[0018] To solve the above technical problems, the present invention also provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the method for determining parameters of an analog circuit as described above are implemented.

[0019] To solve the above technical problems, the present invention also provides a computer program product, including computer program / instructions, and when the computer program / instructions are executed by a processor, the steps of the method for determining parameters of an analog circuit are implemented.

[0020] The beneficial effects of the present invention are as follows. First, based on the circuit source code, the corresponding graph embedding vector is determined to achieve global topology awareness corresponding to the embedding of the circuit topology graph into the graph vector. The graph embedding vector and the performance metrics are input into the policy function, and the local performance feedback and global topology awareness of the performance metric parameters are used as input parameters, significantly improving the pertinence and generalization ability of parameter adjustment. Second, through the policy function, a parameter candidate set for the target analog circuit is obtained. Through multiple sets of sampling processes, the variance of policy gradient estimation is reduced, and the training stability and convergence speed are improved. At the same time, without relying on the value network, the training resource overhead is reduced. Third, multiple parameter combinations of the parameter candidate set are subjected to simulation processing to obtain the corresponding simulation results. The parameter candidate set output by the policy function is applied to the simulation processing scenario to achieve the accuracy of simulation processing evaluation. Fourth, the target performance metrics are extracted from the simulation results to achieve the optimization of the performance metrics. According to the multi-objective reward function, multiple target performance metrics are processed to achieve a balanced trade-off among multiple performance metrics, and the parameter combination corresponding to the highest advantage value is obtained. While improving local and global optimization, the reliability of the process for determining the parameters of the analog circuit is also improved.

[0021] Secondly, the process of determining the graph embedding vector based on the circuit source code facilitates subsequent circuit analysis and the parameter optimization scenario of the analog circuit, and is convenient for data processing. The generation process of the parameter candidate set facilitates the optimization of the policy through within-group comparison, which can reduce the variance of policy updates. Based on the relative performance of multiple samples rather than the absolute performance of a single sample, the policy updates are more stable. It is convenient for subsequent parallel processing and reduces the dependence on independent value functions, thereby reducing the consumption of computing resources. The simulation processing process, through parallel simulation processing, eliminates the need for manual parameter adjustment, greatly shortening the design verification time. It can quickly verify the circuit performance under different parameter configurations, quickly discover potential problems and make adjustments, and reduce the power consumption and cost of the circuit. Based on multiple performance metrics, various analysis methods are used to determine the target performance metrics to improve the fairness and authority of performance metric selection, facilitating subsequent processing of the multi-objective reward function. In the process of constructing the multi-objective reward function, an adaptive weight adjustment mechanism is introduced to achieve a balanced trade-off among multiple target performance metrics. In the process of determining the advantage value, the variance of policy updates is reduced. The role of the baseline is to adjust the reward value so that its deviation from the baseline (i.e., the advantage value) is more stable. This adjustment can reduce the update deviation caused by random fluctuations, thereby improving the optimization efficiency. It makes the policy updates more concentrated on those parameter combinations that are significantly better than the baseline, thereby accelerating the convergence speed of the algorithm. By preferentially updating those parameter combinations with better performance, the algorithm can find the optimal solution faster.

[0022] In addition, the present invention also provides a parameter determination device for an analog circuit, a computer-readable storage medium, and a computer program product, which have the same beneficial effects as the parameter determination method for the analog circuit described above. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] To more clearly illustrate the embodiments of the present invention, the following will briefly introduce the drawings required for use in the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0024] Figure 1 It is a flowchart of a parameter determination method for an analog circuit provided by an embodiment of the present invention; Figure 2 It is a flowchart of the training process of a policy function provided by an embodiment of the present invention; Figure 3 It is a structural diagram of a parameter determination device for an analog circuit provided by an embodiment of the present invention; Figure 4 It is a structural diagram of a parameter determination device for an analog circuit provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0025] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the protection scope of the present invention.

[0026] The core of the present invention is to provide a parameter determination method, device, medium, and product for an analog circuit to solve the technical problems of large training resource overhead, slow convergence speed, and difficulty in optimizing local and global aspects.

[0027] To enable those skilled in the art to better understand the solution of the present invention, the present invention will be further described in detail below in conjunction with the drawings and specific embodiments.

[0028] The parameter design of traditional analog integrated circuits relies on the experience of designers for manual selection and debugging. This process is time-consuming and laborious and highly dependent on design experience. With the development of machine learning technology, in recent years, analog circuit parameter optimization methods based on machine learning have received extensive attention. Existing research usually transforms the circuit parameter optimization problem into a black-box optimization problem and uses methods such as genetic algorithms, Bayesian optimization, particle swarm algorithms, simulated annealing, and deep learning for solution. For example, an analog circuit is regarded as a black-box function from input parameters to a simulator, and the parameters are adjusted according to the simulation output to improve the circuit performance. Traditional reinforcement learning methods have also been used for analog circuit parameter optimization. Traditional reinforcement learning methods rely on a value network in the calculation of the advantage function, which not only increases the training resource overhead but also may lead to estimation errors; its sample efficiency is low in high-dimensional cases and the convergence speed is slow. The method for determining parameters of an analog circuit provided by the present invention can solve the above technical problems.

[0029] Figure 1 The flowchart of a method for determining parameters of an analog circuit provided by an embodiment of the present invention is shown as Figure 1 shown, and the method includes: S11: Obtain the circuit source code and performance indicators corresponding to the target analog circuit; S12: Determine the corresponding graph embedding vector based on the circuit source code, and input the graph embedding vector and the performance indicators into a policy function to obtain a parameter candidate set for the target analog circuit; S13: Perform simulation processing on multiple parameter combinations of the parameter candidate set to obtain corresponding simulation results, and extract corresponding target performance indicators according to the multiple simulation results; S14: Process the multiple target performance indicators according to a multi-objective reward function to obtain corresponding advantage values, and use the parameter combination corresponding to the highest advantage value as the final parameters of the target analog circuit.

[0030] Specifically, the circuit source code in step S11 is code describing the behavior of digital circuits and systems, and the source code for implementing circuit simulation functions can be obtained through an open-source digital circuit simulator or using other programming languages, which is not limited here.

[0031] The performance indicators are quantitative indicators for measuring the performance of a specific system, project, or activity under the target analog circuit, and can help evaluate whether the system or process has reached the expected performance level. There are mainly indicators corresponding to the DC operating point, AC frequency response, transient response, noise analysis, and power consumption. Regarding the number of performance indicators, it can be one or multiple. In order to achieve a balanced trade-off between multiple performance indicators, multiple performance indicators are adopted in this application.

[0032] Determine the graph embedding vector based on the circuit source code in step S12. Use an open-source tool to convert the circuit source code into a netlist, construct a circuit diagram using the netlist, and then map the circuit diagram to a low-dimensional embedding vector to avoid conversion when performing the policy function and match the input format of the policy function.

[0033] Input the graph embedding vector and performance metrics into the policy function, enabling the policy function to not only perceive the physical connections of the circuit but also optimize and adjust in real-time according to the performance feedback. Obtain the parameter candidate set of the target analog circuit. The main feature of the policy function here is a reinforcement learning algorithm that does not rely on a value model. It can be an optimized policy function that directly optimizes behavior selection by adjusting the policy. It can also optimize the policy by learning the action-value function. It can also compare the reward values of multiple actions in the same state to optimize the policy without relying on an independent value network. Regarding the policy function, it can be carried out through the Group Relative Policy Optimization (GRPO) function, and multiple intra-group samplings are performed on each group of parameters to be optimized to obtain the parameter candidate set. It should be noted that there are G different circuit parameter configurations in the parameter candidate set.

[0034] Perform simulation processing on multiple parameter combinations in the parameter candidate set to obtain the corresponding simulation results. Here, the simulation processing is mainly based on the simulation of analog circuits. It can use the simulation software of analog circuits or other programming programs for simulation processing, which is not limited here to obtain the corresponding simulation results. It should be noted here that since there are multiple parameter combinations, the number of simulation results is the same as the number of parameter combinations. Regarding the parameter combination, it is the parameter configuration combination corresponding to multiple components of the target analog circuit.

[0035] Extract the corresponding target performance metrics based on multiple simulation results. Here, the key performance metric is found from multiple performance metrics of the target analog circuit as the target performance metric. The target performance metric corresponds to the specific parameters of the metric.

[0036] The multi-objective reward function in step S14 is a reward function carried out under multiple performance metrics. For each parameter configuration under multiple performance metrics, a reward function is set, and the reward values corresponding to different parameter configurations are obtained through processing. Combining the above G different circuit parameter configurations, they correspond to G reward values. Determine the corresponding advantage values through the reward values, and thus obtain G advantage values. Through the setting of G advantage values, the parameter combination corresponding to the highest advantage value is used as the final parameter of the target analog circuit.

[0037] The beneficial effects of the embodiments of the present invention are as follows. First, the corresponding graph embedding vector is determined based on the circuit source code, realizing the global topology perception corresponding to the embedding of the circuit topology graph into the graph vector. The graph embedding vector and performance metrics are input into the policy function, and the local performance feedback and global topology perception of the performance metric parameters are used as input parameters, significantly improving the pertinence and generalization ability of parameter adjustment. Second, through the policy function, a parameter candidate set of the target analog circuit is obtained. Through multiple groups of sampling processes, the variance of policy gradient estimation is reduced, and the training stability and convergence speed are improved. At the same time, without relying on a value network, the training resource overhead is reduced. Third, multiple parameter combinations of the parameter candidate set are simulated to obtain corresponding simulation results. The parameter candidate set output by the policy function is applied to the simulation processing scenario, realizing the accuracy of simulation processing evaluation. Fourth, the target performance metrics are extracted from the simulation results to optimize the performance metrics. According to the multi-objective reward function, multiple target performance metrics are processed to achieve a balanced trade-off among multiple performance metrics, and the parameter combination corresponding to the highest advantage value is obtained, improving both local and global optimization and also the reliability of the process for determining the parameters of the analog circuit.

[0038] In some embodiments, determining the corresponding graph embedding vector based on the circuit source code includes: Converting the circuit source code into a netlist; Constructing a circuit diagram of the target analog circuit based on the netlist; Invoking a graph neural network model, inputting the circuit diagram into the graph neural network model to obtain the corresponding graph embedding vector.

[0039] Specifically, when converting the circuit source code into a netlist, the circuit source code usually exists in the form of a hardware description language. To convert the circuit source code into a netlist, a logic synthesis tool or an open-source tool is used for conversion. To convert the netlist into a visual circuit diagram, a conversion tool can be used. The purpose of the conversion is to make the circuit diagram structure processable. Constructing a circuit diagram of the target analog circuit based on the netlist facilitates subsequent analysis and optimization.

[0040] Inputting the circuit diagram into the graph neural network model to obtain the corresponding graph embedding vector. The specific way of generating the graph embedding vector can be the same as or different from the conventional model generation method, and can be set according to the actual situation here.

[0041] The process of determining the graph embedding vector based on the circuit source code provided in this embodiment facilitates subsequent circuit analysis and parameter optimization scenarios of analog circuits and is convenient for data processing.

[0042] In some embodiments, constructing a circuit diagram of the target analog circuit based on the netlist includes: Extract the netlist nodes and netlist edge information corresponding to the netlist from the netlist, where the netlist nodes represent the components of the target analog circuit, and the netlist edge information represents the connection relationship between the components; Process the netlist nodes and netlist edge information through a graph data structure to obtain a circuit diagram.

[0043] Specifically, a netlist is a text file describing the circuit connection relationship, mainly composed of netlist nodes, netlist edge information, and ports. Netlist nodes are components in the circuit, such as resistors, capacitors, and transistors. Netlist edge information is the connection relationship between components. Ports are input and output ports. The netlist file represents the circuit connection relationship in a specific format. Component declarations are the types, names, and connection nodes of each component.

[0044] The netlist nodes and netlist edge information extracted from the netlist are used to represent the circuit through a graph data structure. The graph can be represented using an adjacency list or an adjacency matrix.

[0045] The circuit diagram constructed from the netlist in this embodiment can visually view the circuit topology diagram, the connection relationship and layout of multiple components, laying a foundation for subsequent generation of graph embedding vectors.

[0046] In some embodiments, the process of determining the parameter candidate set includes: Pre-establish a parameter database for the target analog circuit; where the parameter database stores the parameters to be determined for the target analog circuit and their corresponding value ranges; Determine the value range of the output neuron output value of the policy function according to the parameter database; Perform multiple random samplings on the output value of the policy function according to the graph embedding vector and performance metrics to generate a parameter candidate set.

[0047] Specifically, the parameter database of the target analog circuit stores the parameters to be determined for the target analog circuit. The parameters to be determined only know the types of parameters and the value ranges of the corresponding specific parameters. Regarding the parameter configuration combination, it can be determined in a combined set manner or a list manner through a mapping relationship, and can be set according to the actual situation.

[0048] Use the parameter database as the function parameter of the policy function. Here, the function parameter corresponds to the action space of the policy function, so that the specific value of the parameter is restricted not to exceed this value range. Perform multiple random samplings on the embedding vector and performance metrics to generate a parameter candidate set, that is, G different circuit parameter configurations.

[0049] The generation process of the parameter candidate set provided in this embodiment, which facilitates optimizing the strategy through within-group comparison, can reduce the variance of strategy updates. Based on the relative performance of multiple samples rather than the absolute performance of a single sample, it makes the strategy updates more stable. It is convenient for subsequent parallel processing, reduces the dependence on independent value functions, and thus reduces the consumption of computing resources.

[0050] In some embodiments, simulating multiple parameter combinations of the parameter candidate set to obtain corresponding simulation results includes: Sequentially applying multiple parameter combinations to the netlist; Invoking a simulator to perform circuit simulation on the netlist to obtain corresponding simulation results.

[0051] Specifically, when simulating multiple parameter combinations, the number of corresponding simulation results is the same as the number of parameter combinations. For multiple parameter combinations, they are sequentially applied to the netlist, and a simulator is invoked to perform circuit simulation to obtain corresponding simulation results.

[0052] The simulator here can be an Integrated Circuit Emphasis (SPICE) simulator or other simulators, which is not limited herein. However, during the simulation process, in order for the obtained simulation results to follow the same simulation mechanism, the same simulator is used.

[0053] The simulation process provided in this embodiment, through parallel simulation, eliminates the need for manual parameter adjustment, greatly shortens the design verification time, can quickly verify the circuit performance under different parameter configurations, can quickly discover potential problems and make adjustments, and reduces the power consumption and cost of the circuit.

[0054] In some embodiments, extracting corresponding target performance indicators according to multiple simulation results includes: Determining the common performance indicator corresponding to the frequency response according to the functional requirements, performance requirements, multiple parameter configurations, and sensitivity analysis of the target analog circuit among the performance indicators of the target analog circuit; Taking the common performance indicator as the target performance indicator; Determining the corresponding target parameters according to the target performance indicator among multiple simulation results.

[0055] Specifically, there are quite a few performance indicators involved in the analog circuit. How to find the key performance indicators among them can be to select those that can comprehensively reflect the circuit performance and are closely related to the design goal.

[0056] Specifically, the target performance indicators can be determined by any one of the analysis methods such as functional requirement analysis, performance requirement analysis, multiple parameter configuration combinations, and sensitivity analysis. It is also possible to comprehensively determine the common performance indicators by combining multiple analysis methods as the target performance indicators.

[0057] Functional requirement analysis mainly corresponds to components such as amplifiers, filters, and oscillators. The performance indicators corresponding to an amplifier are gain, bandwidth, input impedance, and output impedance; the performance indicators corresponding to a filter are cut-off frequency, passband gain, and stopband attenuation; the performance indicators corresponding to an oscillator are oscillation frequency, phase noise, and harmonic distortion. Performance requirement analysis can be low-noise requirements (noise power, signal-to-noise ratio), high-bandwidth requirements (bandwidth, gain flatness), and low-power requirements (static power consumption, dynamic power consumption). Combinations of multiple parameter configurations include different component values (the gain and bandwidth of the circuit at different resistor and capacitor values), different power supply voltages (the power consumption and performance of the circuit at different power supply voltages), and different temperature ranges (the noise and stability of the circuit at different temperatures). Sensitivity analysis is to determine which performance indicators are most sensitive to parameter changes, such as the sensitivity of gain to resistor values: analyzing the change of gain with resistor values through simulation; the sensitivity of bandwidth to capacitor values: analyzing the change of bandwidth with capacitor values through simulation.

[0058] The common performance indicators are the performance indicators that need to be considered in the above different analysis methods, such as circuit gain, bandwidth, power consumption, and linearity indicators. The linearity indicator is an important indicator to measure the degree of linear relationship between the output and input of a sensor or measurement system. Based on these indicators, the corresponding target parameter values, that is, reference values, are determined under this indicator.

[0059] This embodiment provides a method for determining target performance indicators by using multiple analysis methods among multiple performance indicators to improve the fairness and authority of performance indicator selection and facilitate subsequent processing of multi-objective reward functions.

[0060] In some embodiments, multiple target performance indicators are processed according to a multi-objective reward function to obtain corresponding advantage values, including: Obtaining weight parameters corresponding to multiple target performance indicators; Substituting multiple target performance indicators, weight parameters, and target parameters into the multi-objective reward function to obtain reward values corresponding to multiple parameter combinations; Determining the average value and variance value of the reward values of multiple parameter combinations; Using the average value and variance value as the baseline to perform advantage processing on multiple parameter combinations to obtain corresponding advantage values.

[0061] Specifically, obtain the weight parameters corresponding to the target performance metrics, and substitute the target performance metrics, the corresponding weight parameters, and the target parameters into the multi-objective reward function to obtain the reward values corresponding to multiple parameter combinations.

[0062] Use the average value and variance value of the reward values of multiple parameter combinations as the baseline, and perform advantage processing on each parameter combination to obtain the corresponding advantage value.

[0063] The processing process provided in this embodiment for processing multiple target performance metrics according to the multi-objective reward function to obtain the corresponding advantage value does not require an additional value network, reducing the training overhead cost.

[0064] In some embodiments, the target performance metrics are at least a circuit gain metric, a bandwidth metric, a power consumption metric, and a linearity metric; substituting multiple target performance metrics, weight parameters, and target parameters into the multi-objective reward function to obtain the reward values corresponding to multiple parameter combinations includes: Obtain the first weight parameter corresponding to the circuit gain metric, the second weight parameter corresponding to the bandwidth metric, the third weight parameter corresponding to the power consumption metric, and the fourth weight parameter corresponding to the linearity metric; Determine the first ratio value according to the circuit gain metric and the corresponding target parameter; Determine the second ratio value according to the bandwidth metric and the corresponding target parameter; Determine the third ratio value according to the power consumption metric and the corresponding target parameter; Determine the fourth ratio value according to the linearity metric and the corresponding target parameter; Process the first ratio value and the first weight parameter, the second ratio value and the second weight parameter, the third ratio value and the third weight parameter, and the fourth ratio value and the fourth weight parameter to obtain the reward values corresponding to multiple parameter combinations.

[0065] Specifically, set the corresponding weights for the four target performance metrics respectively, and determine the corresponding ratio values for different target performance metrics and their respective specific parameters. Process based on the ratio values and weight parameters to obtain the reward values.

[0066] The process of constructing the multi-objective reward function provided in this embodiment introduces an adaptive weight adjustment mechanism, enabling a balanced trade-off among multiple target performance metrics.

[0067] In some embodiments, processing the first ratio value and the first weight parameter, the second ratio value and the second weight parameter, the third ratio value and the third weight parameter, and the fourth ratio value and the fourth weight parameter to obtain the reward values corresponding to multiple parameter combinations includes: Process the first ratio value and the first weight parameter to obtain the first reward value; Process the second proportional value and the second weight parameter to obtain a second reward value; Process the third proportional value and the third weight parameter to obtain a third reward value; Process the fourth proportional value and the fourth weight parameter to obtain a fourth reward value; Add the first reward value and the second reward value to obtain a fifth reward value; Subtract the third reward value and the fourth reward value from the fifth reward value in sequence to obtain the final reward value.

[0068] Specifically, the formula is as follows: ; Wherein, , , , are the circuit gain index, bandwidth index, power consumption index, and linearity index respectively; , , , are the target parameters corresponding to the circuit gain index, bandwidth index, power consumption index, and linearity index respectively; , , , correspond to the first proportional value, second proportional value, third proportional value, and fourth proportional value respectively. , , and correspond to the first weight parameter, second weight parameter, third weight parameter, and fourth weight parameter respectively.

[0069] The construction process of the multi-objective reward function provided in this embodiment is based on the weighting of different target performance indicators, realizing a flexible balance among the target performance indicators of gain, bandwidth, power consumption, and linearity.

[0070] In some embodiments, the sum of the first weight parameter and the second weight parameter is 1; the sum of the third weight parameter and the fourth weight parameter is 1.

[0071] Combined with the above embodiments, after the weighted summation of the first weight parameter and the second weight parameter, it is necessary to subtract the weighted summation corresponding to the third weight parameter and the fourth weight parameter, which is considered in view of the consumption corresponding to the power consumption and linearity indicators in the actual hardware circuit. While optimizing multiple performance indicators, subtracting power consumption and linearity is to ensure finding a balance between performance and resources during the optimization process, and to avoid over-optimizing a certain indicator and thus ignoring other important indicators.

[0072] In some embodiments, the training process of the policy function includes: Call the initial policy function and input the current performance metric and the graph embedding vector into the initial policy function; wherein, the current performance metric is the previous target performance metric obtained during the previous training of the policy function, and is the initial performance metric during the first training. Output the current parameter candidate set corresponding to the initial policy function. Perform simulation processing on multiple parameter combinations of the current parameter candidate set to obtain the corresponding current simulation results, so as to extract the corresponding current target performance metrics. Process multiple current target performance metrics according to the multi-objective reward function to obtain the corresponding current advantage values. Perform fine-tuning processing on the initial policy function based on the current advantage values to obtain the policy function after parameter fine-tuning processing. When the current iteration number reaches the preset iteration number, then use the initial policy function as the final policy function. When the current iteration number does not reach the preset iteration number, then use the policy function after parameter fine-tuning processing as the new policy function, and return to the step of inputting the current performance metric and the graph embedding vector into the initial policy function, until the current iteration number reaches the preset iteration number, and determine the final policy function.

[0073] Specifically, in the reinforcement learning environment, define the state space, action space, and reward function. The state can be composed of the performance metrics of the circuit (such as the current gain, bandwidth, etc.), and the action corresponds to the value of the parameter to be determined. The input layer of the policy network receives the embedding vector + performance metric corresponding to the circuit; the hidden layer uses a multi-layer fully connected or structured attention mechanism; the output layer predicts the circuit parameter adjustment action (such as the bias voltage). The initial state can be randomly valued. For example, the weights of the policy network are randomly initialized and the bias terms are set to zero.

[0074] Use open-source tools to convert the circuit source code into a netlist, use the netlist nodes as graph nodes and the connections as graph edges to construct a circuit diagram, and then use a graph neural network to map the circuit diagram to a low-dimensional embedding vector.

[0075] Embed the figure corresponding to the circuit diagram into a vector and use the performance metrics obtained in the previous round as inputs. Randomly sample a set of current parameter candidate sets (i.e., A different circuit parameter configurations) according to the current initial policy function; apply the parameter values to the circuit netlist and call the SPICE simulator to perform circuit simulation to obtain the corresponding simulation results; extract the key performance metrics of the circuit from the simulation results, and use a pre-designed multi-objective reward function to calculate the reward value for each parameter configuration. Calculate the advantage value for each parameter configuration based on the mean and variance values of the reward values as a baseline. Fine-tune the parameters of the initial policy function according to the advantage value to determine a new policy function, and record the current iteration number. When the current iteration number reaches the preset iteration number, it indicates that the initial policy function has been trained. If the current iteration number has not reached the preset iteration number, use the policy function after parameter fine-tuning as the new policy function and continue training until the current iteration number reaches the preset iteration number, at which point the training is complete and the final policy function is determined.

[0076] In addition to the constraint of the iteration number, it can also be determined that the performance metrics meet the requirements of the preset metrics.

[0077] The training process of the policy function provided in this embodiment can automatically parse the circuit netlist, construct a circuit topology diagram and extract the graph embedding through a graph neural network. At the same time, performance metrics such as gain, bandwidth, power consumption, and linearity obtained from the previous round of SPICE simulation are used as auxiliary inputs, enabling the policy network to not only perceive the physical connection of the circuit but also perform optimization and adjustment in real time according to the performance feedback. After sending the multi-modal input into the policy function, the network performs multiple intra-group samplings on each group of actions and calculates the average reward baseline to achieve a robust search in the high-dimensional parameter space. This method significantly reduces the variance of the policy gradient estimation, improves the training stability and convergence speed, and is more suitable for the global optimization of complex analog circuits compared to other algorithms. Construct a composite reward function and introduce an adaptive weight adjustment mechanism to flexibly balance and trade-off between objectives such as gain, bandwidth, power consumption, and linearity.

[0078] In some embodiments, fine-tuning the initial policy function based on the advantage value to obtain the policy function after parameter fine-tuning includes: Substitute the advantage value into the loss function of the initial policy function, and perform gradient processing on the loss function to obtain the adjustment amounts of the weights and biases of the policy function; Apply the adjustment amounts of the weights and biases to the initial policy function to complete the parameter fine-tuning process of the initial policy function.

[0079] Specifically, by optimizing the policy function, calculating the gradient and updating the current policy function, perform gradient processing based on the advantage value. The specific processing method can be the same as or different from the conventional gradient descent method, and can be set according to the actual situation.

[0080] The process of the optimization strategy function provided in this embodiment can accelerate the optimization process of parameter weights and bias terms, effectively avoid falling into local optimal solutions, adjust the learning rate and the number of iterations, ensure searching for the optimal solution within the global scope, and improve the optimization quality.

[0081] In some embodiments, taking the average value and the variance value as baselines, performing an advantage process on multiple parameter combinations to obtain corresponding advantage values, includes: Performing a subtraction process on the current reward value and the average value to obtain a sixth reward value; Performing a division process on the square root of the variance value and the sixth reward value to obtain the corresponding advantage value.

[0082] Regarding the advantage calculation process, the specific formula is as follows: ; Wherein, is the current reward value, is the average value, is the sixth reward value, is the square root of the variance value, represents the th parameter combination.

[0083] The process of determining the advantage value provided in this embodiment reduces the variance of policy updates. The role of the baseline is to adjust the reward value to make its deviation relative to the baseline (i.e., the advantage value) more stable. This adjustment can reduce the update deviation caused by random fluctuations, thereby improving the optimization efficiency. It makes the policy updates more concentrated on those parameter combinations that are significantly better than the baseline, thus accelerating the convergence speed of the algorithm. By preferentially updating those parameter combinations with better performance, the algorithm can find the optimal solution faster.

[0084] Figure 2 is a flowchart of the training process of a policy function provided by an embodiment of the present invention. As Figure 2 shown, it includes: S21: According to the circuit structure and design requirements, determine the parameters to be determined and their value ranges; S22: Convert the circuit source code into a netlist, construct a circuit diagram, and then use a graph neural network to obtain a graph embedding vector; S23: Initialize the weights and biases of the policy function; S24: Use the graph embedding vector and the performance metrics obtained in the previous round as inputs, so that the policy function generates a set of parameter candidate sets through random sampling; S25: Apply the parameter candidate sets to the circuit netlist respectively, and call the simulator to obtain the simulation results; S26: Extract the performance metrics from the simulation results, and calculate the reward value and the corresponding advantage value; S27: Update the weights and biases of the policy function using gradient calculation based on the loss function and the advantage value; S30: Determine whether the number of training rounds has reached a preset threshold. If so, end; if not, return to step S24.

[0085] The above has described in detail each embodiment corresponding to the method for determining the parameters of the analog circuit. On this basis, the present invention also discloses a device for determining the parameters of the analog circuit corresponding to the above method. Figure 3 It is a structural diagram of a device for determining the parameters of an analog circuit provided by an embodiment of the present invention. As Figure 3 shown, the device for determining the parameters of the analog circuit includes: An acquisition module 11, configured to acquire the circuit source code and performance indicators corresponding to the target analog circuit; An input module 12, configured to determine the corresponding graph embedding vector based on the circuit source code, and input the graph embedding vector and the performance indicators into the policy function to obtain a parameter candidate set of the target analog circuit; An extraction module 13, configured to perform simulation processing on multiple parameter combinations of the parameter candidate set to obtain corresponding simulation results, and extract corresponding target performance indicators according to the multiple simulation results; A processing module 14, configured to process the multiple target performance indicators according to the multi-objective reward function to obtain corresponding advantage values, and use the parameter combination corresponding to the highest advantage value as the final parameters of the target analog circuit.

[0086] Since the embodiments of the device part correspond to the above embodiments, the embodiments of the device part are described with reference to the embodiments of the above method part and will not be elaborated here.

[0087] For the introduction of a device for determining the parameters of an analog circuit provided by the present invention, please refer to the above method embodiments. The present invention will not elaborate here, and it has the same beneficial effects as the above method for determining the parameters of the analog circuit.

[0088] Figure 4 It is a structural diagram of a device for determining the parameters of an analog circuit provided by an embodiment of the present invention. As Figure 4 shown, the device includes: A memory 21, configured to store a computer program; A processor 22, configured to implement the steps of the method for determining the parameters of the analog circuit when executing the computer program.

[0089] The device for determining the parameters of the analog circuit provided in this embodiment may include but is not limited to a smart phone, a tablet computer, a notebook computer, or a desktop computer, etc.

[0090] Among them, the processor 22 may include one or more processing cores, such as a 4-core processor, an 8-core processor, etc. The processor 22 may be implemented in at least one hardware form of a Digital Signal Processor (DSP), a Field-Programmable Gate Array (FPGA), and a Programmable Logic Array. The processor 22 may also include a main processor and a coprocessor. The main processor is a processor used to process data in the wake state, also known as the Central Processing Unit (CPU); the coprocessor is a low-power processor used to process data in the standby state. In some embodiments, the processor 22 may be integrated with a Graphics Processing Unit (GPU), and the GPU is responsible for rendering and drawing the content to be displayed on the display screen. In some embodiments, the processor 22 may further include an Artificial Intelligence (AI) processor, and the AI processor is used to process computational operations related to machine learning.

[0091] The memory 21 may include one or more computer-readable storage media, and the computer-readable storage media may be non-transitory. The memory 21 may further include high-speed random access memory and non-volatile memory, such as one or more disk storage devices and flash storage devices. In this embodiment, the memory 21 is at least used to store the following computer program 211. After the computer program is loaded and executed by the processor 22, it can implement the relevant steps of the parameter determination method of the analog circuit disclosed in any of the foregoing embodiments. In addition, the resources stored in the memory 21 may also include an operating system 212 and data 213, etc., and the storage method may be temporary storage or permanent storage. Among them, the operating system 212 may include Windows, Unix, Linux, etc. The data 213 may include, but is not limited to, the data involved in the parameter determination method of the analog circuit, etc.

[0092] In some embodiments, the analog circuit parameter determination device may further include a display screen 23, an input / output interface 24, a communication interface 25, a power supply 26, and a communication bus 27.

[0093] Those skilled in the art can understand that Figure 4 the structure shown in

[0094] does not constitute a limitation on the analog circuit parameter determination device, and may include more or fewer components than shown in the figure.

[0095] For the introduction of a parameter determination device for an analog circuit provided by the present invention, please refer to the above method embodiments. The present invention will not repeat it here, and it has the same beneficial effects as the above parameter determination method for the analog circuit.

[0096] Furthermore, the present invention also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor 22, the steps of the parameter determination method for the analog circuit as described above are implemented.

[0097] It can be understood that if the methods in the above embodiments are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and executes all or part of the steps of the methods in various embodiments of the present invention. The foregoing storage media include: various media such as USB flash drives, mobile hard disks, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical discs that can store program codes.

[0098] For the introduction of a computer-readable storage medium provided by the present invention, please refer to the above method embodiments. The present invention will not repeat it here, and it has the same beneficial effects as the above parameter determination method for the analog circuit.

[0099] Further, the present invention also provides a computer program product, including computer programs / instructions. When the computer programs / instructions are executed by a processor, the steps of the parameter determination method for the analog circuit are implemented.

[0100] For the introduction of a computer program product provided by the present invention, please refer to the above method embodiments. The present invention will not repeat it here, and it has the same beneficial effects as the above parameter determination method for the analog circuit.

[0101] The above has introduced in detail a method, device, medium, and product for determining parameters of an analog circuit provided by the present invention. The various embodiments in the specification are described in a progressive manner. Each embodiment focuses on the differences from other embodiments. For the same or similar parts among the various embodiments, reference can be made to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and for the relevant parts, reference can be made to the description in the method part. It should be noted that for those of ordinary skill in the art of the present technology, without departing from the principle of the present invention, several improvements and modifications can be made to the present invention, and these improvements and modifications also fall within the protection scope of the present invention.

[0102] It should also be noted that in this specification, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including", or any other variant thereof is intended to cover non-exclusive inclusion, such that a process, method, article, or device comprising a series of elements includes not only those elements but also other elements not explicitly listed, or further includes elements inherent to such process, method, article, or device. Without further limitation, an element defined by the statement "comprising one..." does not exclude the existence of another identical element in the process, method, article, or device comprising the element.

Claims

1. A method for determining parameters of an analog circuit, characterized in that Including: Obtain the circuit source code and performance metrics corresponding to the target analog circuit; Based on the circuit source code, determine the corresponding graph embedding vector, and input the graph embedding vector and the performance metrics into a policy function to obtain a parameter candidate set for the target analog circuit; Perform simulation processing on multiple parameter combinations in the parameter candidate set to obtain corresponding simulation results, and extract corresponding target performance metrics according to the multiple simulation results; Process the multiple target performance metrics according to a multi-objective reward function to obtain corresponding advantage values, and use the parameter combination corresponding to the highest advantage value as the final parameters of the target analog circuit.

2. The method for determining parameters of an analog circuit according to claim 1, wherein The determination process of the parameter candidate set includes: Pre-establish a parameter database for the target analog circuit; wherein, the parameter database stores the parameters to be determined for the target analog circuit and their corresponding value ranges; Determine the value range of the output neuron output value of the policy function according to the parameter database; Perform multiple random samplings on the output value of the policy function according to the graph embedding vector and the performance metrics to generate the parameter candidate set.

3. The method for determining the parameters of the analog circuit according to claim 1, wherein Performing simulation processing on multiple parameter combinations in the parameter candidate set to obtain corresponding simulation results includes: Apply multiple parameter combinations to the netlist in sequence; Call a simulator to perform circuit simulation processing on the netlist to obtain corresponding simulation results.

4. The method for determining the parameters of the analog circuit according to claim 3, characterized in that, Extracting corresponding target performance metrics according to multiple simulation results includes: Determine the common performance metric corresponding to the frequency response according to the functional requirements, performance requirements, multiple parameter configurations, and sensitivity analysis of the circuit among the performance metrics of the target analog circuit; Use the common performance metric as the target performance metric; Determine the corresponding target parameters according to the target performance metric among the multiple simulation results.

5. The method for determining the parameters of the analog circuit according to claim 4, wherein Processing the multiple target performance metrics according to a multi-objective reward function to obtain corresponding advantage values includes: Obtain the weight parameters corresponding to the multiple target performance metrics; Substitute the multiple target performance metrics, weight parameters, and target parameters into the multi-objective reward function to obtain the reward values corresponding to multiple parameter combinations; Determine the average value and variance value of the reward values of multiple parameter combinations; Use the average value and the variance value as a baseline to perform advantage processing on multiple parameter combinations to obtain corresponding advantage values.

6. The method for determining parameters of an analog circuit according to claim 5, characterized in that, The target performance metrics are at least circuit gain metrics, bandwidth metrics, power consumption metrics, and linearity metrics; substituting the multiple target performance metrics, weight parameters, and target parameters into the multi-objective reward function to obtain the reward values corresponding to multiple parameter combinations includes: Obtain the first weight parameter corresponding to the circuit gain metric, the second weight parameter corresponding to the bandwidth metric, the third weight parameter corresponding to the power consumption metric, and the fourth weight parameter corresponding to the linearity metric; Determine the first ratio value according to the circuit gain metric and the corresponding target parameter; Determine the second ratio value according to the bandwidth metric and the corresponding target parameter; Determine the third ratio value according to the power consumption metric and the corresponding target parameter; Determine the fourth ratio value according to the linearity metric and the corresponding target parameter; Process the first ratio value and the first weight parameter, the second ratio value and the second weight parameter, the third ratio value and the third weight parameter, and the fourth ratio value and the fourth weight parameter to obtain the reward values corresponding to multiple parameter combinations.

7. The method for determining the parameters of the analog circuit according to claim 6, wherein Processing the first ratio value and the first weight parameter, the second ratio value and the second weight parameter, the third ratio value and the third weight parameter, and the fourth ratio value and the fourth weight parameter to obtain the reward values corresponding to multiple parameter combinations includes: Processing the first ratio value and the first weight parameter to obtain a first reward value; Processing the second ratio value and the second weight parameter to obtain a second reward value; Processing the third ratio value and the third weight parameter to obtain a third reward value; Processing the fourth ratio value and the fourth weight parameter to obtain a fourth reward value; Adding the first reward value and the second reward value to obtain a fifth reward value; Subtracting the third reward value and the fourth reward value from the fifth reward value in sequence to obtain the final reward value.

8. The method for determining parameters of the analog circuit according to claim 6, characterized in that, The sum of the first weight parameter and the second weight parameter is 1; the sum of the third weight parameter and the fourth weight parameter is 1.

9. The method for determining parameters of an analog circuit according to claim 5, characterized in that, The training process of the policy function includes: Call the initial policy function and input the current performance metric and the graph embedding vector into the initial policy function; where the current performance metric is the previous target performance metric obtained during the previous training of the policy function and is the initial performance metric during the first training; Output the current parameter candidate set corresponding to the initial policy function; Perform simulation processing on multiple parameter combinations of the current parameter candidate set to obtain the corresponding current simulation results, so as to extract the corresponding current target performance metrics; Process multiple current target performance metrics according to the multi-objective reward function to obtain the corresponding current advantage values; Perform fine-tuning processing on the initial policy function based on the current advantage values to obtain the policy function after parameter fine-tuning processing; When the current iteration number reaches the preset iteration number, then use the initial policy function as the final policy function; When the current iteration number does not reach the preset iteration number, then use the policy function after parameter fine-tuning processing as the new policy function, and return to the step of inputting the current performance metric and the graph embedding vector into the initial policy function until the current iteration number reaches the preset iteration number, and determine the final policy function.

10. The method for determining the parameters of the analog circuit according to claim 5, wherein Using the average value and the variance value as the baseline, performing advantage processing on multiple parameter combinations to obtain the corresponding advantage values includes: Subtracting the current reward value from the average value to obtain a sixth reward value; Dividing the square root of the variance value by the sixth reward value to obtain the corresponding advantage value.

11. The method for determining the parameters of the analog circuit according to claim 1, characterized in that, Determining the corresponding graph embedding vector based on the circuit source code includes: Converting the circuit source code into a netlist; Constructing the circuit diagram of the target analog circuit based on the netlist; Calling the graph neural network model and inputting the circuit diagram into the graph neural network model to obtain the corresponding graph embedding vector.

12. The method for determining the parameters of the analog circuit according to claim 11, characterized in that, Constructing the circuit diagram of the target analog circuit based on the netlist includes: Extract the netlist nodes and netlist edge information corresponding to the netlist from the netlist, where the netlist nodes represent the components of the target analog circuit, and the netlist edge information represents the connection relationship between the components; Process the netlist nodes and the netlist edge information through a graph data structure to obtain the circuit diagram.

13. A parameter determination device for an analog circuit, characterized in that, Comprising: A memory for storing a computer program; A processor for implementing the steps of the method for determining parameters of an analog circuit according to any one of claims 1 to 12 when executing the computer program.

14. A computer-readable storage medium, characterized in that, A computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, the steps of the method for determining parameters of an analog circuit according to any one of claims 1 to 12 are implemented.

15. A computer program product, comprising a computer program / instructions, characterized in that, When the computer program / instructions are executed by a processor, the steps of the method for determining parameters of an analog circuit according to any one of claims 1 to 11 are implemented.

Images