An automatic parameter adjustment system and method for pre-simulation process of operational amplifier design

Abstract

The present application is an automatic parameter adjustment system and method for the pre-simulation process of operational amplifier design, including: a drawing input module for drawing the operational amplifier schematic diagram, a simulation test module for taking the open-loop gain value as the target, reading the drawn operational amplifier schematic diagram, performing simulation test and outputting the simulation test results; a reinforcement learning module for reading the operational amplifier performance parameters obtained by the simulation test, automatically adjusting the parameter values ​​of each device in the operational amplifier that needs to be adjusted, controlling the cyclic iterative test of the simulation test module, and comparing and judging with the open-loop gain value until the open-loop gain value is met; an output storage module for storing the simulation test results and the operational amplifier schematic diagram after parameter adjustment; in the actual application process, the processing optimization of the simulation test process and the adjustment of the parameters of each device are effectively integrated, greatly improving the efficiency of the pre-simulation parameter adjustment work of the operational amplifier design and the reliability of the optimization scheme.

Description

Technical Field

[0001] The present application relates to the field of analog integrated circuit design, and in particular to an automatic parameter adjustment system and method for a pre-simulation process of operational amplifier design. Background Art

[0002] An operational amplifier is a circuit unit with a very high amplification factor. In actual circuits, it is usually combined with a feedback network to form a certain functional module. The operational amplifier design process has high requirements for performance and efficiency, so it requires high design experience from the designer. In the pre-simulation process of the operational amplifier design, the designer needs to continuously simulate the operational amplifier schematic and adjust the parameter values ​​of each device in the schematic in the shortest possible time until the performance parameters of the operational amplifier pre-simulation meet the open-loop gain value of the operational amplifier design. The design process is extremely time-consuming and labor-intensive.

[0003] The traditional operational amplifier design method requires that after the operational amplifier schematic is drawn, simulation tests are performed to obtain the current performance parameters of the operational amplifier, and then the parameter values ​​of each device are manually adjusted based on the simulation results. Then, the schematic diagram is re-simulated to verify whether the manually modified schematic diagram can meet the requirements of the operational amplifier design open-loop gain value. The above process is repeated until the more ideal parameter values ​​of each device are found.

[0004] The traditional method of pre-simulation parameter adjustment for operational amplifier design requires a lot of pre-simulation preparation work to be completed in the software, which consumes a lot of energy. At the same time, due to the low efficiency and unclear direction of manual operation, several cycles of parameter adjustment are required. Due to the cumbersomeness of verification, the development speed is slow in actual operation, which is not conducive to the efficient implementation of the operational amplifier design process. Summary of the invention

[0005] In order to solve the problems in the prior art that operational amplifier design requires simulation testing, manual operation consumes a lot of energy, is inefficient, and the design process is slow, the present application provides an automatic parameter adjustment system for the pre-simulation process of operational amplifier design, including: a drawing input module, a simulation test module, a reinforcement learning module, and an output storage module;

[0006] The drawing input module is used to draw the operational amplifier schematic diagram, and the simulation test module is used to read the drawn operational amplifier schematic diagram with the open-loop gain value as the target, perform simulation test and output the simulation test result; the simulation test result is the operational amplifier performance parameter, and the operational amplifier performance parameter corresponds to the open-loop gain value;

[0007] The reinforcement learning module is used to read the simulation test result and determine whether the simulation test result meets the open-loop gain value;

[0008] If the simulation test result completely complies with the open-loop gain value, the simulation test result and the schematic diagram are output to the output storage module, and the storage module is used to store the simulation test result and the operational amplifier schematic diagram after parameter adjustment;

[0009] If the simulation test result does not meet the open-loop gain value, the reinforcement learning module adds a penalty value to the simulation test result according to the zero-order optimization algorithm, the covariance matrix adaptive evolution strategy algorithm, the Bayesian optimization algorithm and the expert prior algorithm, automatically generates new parameters of each device of the operational amplifier, modifies the parameters of each device in the schematic diagram, and outputs the modified schematic diagram to the simulation test module for re-simulation test. The above process is repeated until the simulation test result meets the open-loop gain value, and the schematic diagram and the simulation test result are stored.

[0010] Furthermore, the expert a priori algorithm includes a penalty function, which is obtained by calculating an open-loop gain value and an operational amplifier performance parameter obtained by simulation testing.

[0011] Furthermore, the penalty function is specifically:

[0012]

[0013]

[0014] Where: f is the return value of the penalty function, f i is one of the penalty function summation formulas, N is the number of open-loop gain values, x i are the open-loop gain values, is the performance parameter corresponding to each open-loop gain value.

[0015] Furthermore, the penalty function is:

[0016]

[0017] Where: f i is one of the penalty function summation formulas, x i is the open-loop gain value with the minimum value in the penalty function, is the performance parameter corresponding to the open-loop gain value.

[0018] Furthermore, the operational amplifier schematic diagram also includes: component graphic symbols, component text symbols, component netlist, component coordinates, component rotation angle, coordinates of each pin of the component, connecting lines, component bit number coordinates, circuit node coordinates and text annotations.

[0019] Furthermore, the above steps are all completed under the Linux computer system.

[0020] An automatic parameter adjustment method for a pre-simulation process of an operational amplifier design is applied to an automatic parameter adjustment system for a pre-simulation process of an operational amplifier design as described above. The automatic parameter adjustment method for operational amplifier design comprises:

[0021] Draw an operational amplifier schematic diagram, use the open-loop gain value as a target, read the drawn operational amplifier schematic diagram, perform simulation tests and output simulation test results;

[0022] The drawing input module is used to draw the operational amplifier schematic diagram, and is used to read the drawn operational amplifier schematic diagram with the open-loop gain value as the target, perform simulation test and output the simulation test result;

[0023] Reading the simulation test result, and determining whether the simulation test result meets the open-loop gain value,

[0024] If the simulation test result completely complies with the open-loop gain value, the simulation test result and the schematic diagram are output to the output storage module, and the storage module is used to store the simulation test result and the operational amplifier schematic diagram after parameter adjustment;

[0025] If the simulation test result does not meet the open-loop gain value, the reinforcement learning module adds a penalty value to the simulation test result according to the zero-order optimization algorithm, the covariance matrix adaptive evolution strategy algorithm, the Bayesian optimization algorithm and the expert prior algorithm, automatically generates new parameters of each device of the operational amplifier, modifies the parameters of each device in the schematic diagram, and outputs the modified schematic diagram to the simulation test module for re-simulation test. The above process is repeated cyclically until the simulation test result meets the open-loop gain value, and the schematic diagram and the simulation test result are stored.

[0026] A computer device comprising:

[0027] Memory for storing computer programs;

[0028] A processor is used to implement the steps of the automatic parameter adjustment method based on an operational amplifier design pre-simulation process as described above when executing the computer program.

[0029] From the above technical scheme, it can be known that an automatic parameter adjustment system for the pre-simulation process of operational amplifier design includes: a drawing input module, a simulation test module, a reinforcement learning module and an output storage module; the drawing input module is used to draw the operational amplifier schematic diagram, the simulation test module is used to read the drawn operational amplifier schematic diagram with the open-loop gain value as the target, perform simulation test and output the simulation test result; the simulation test result is the operational amplifier performance parameter, and the operational amplifier performance parameter corresponds to the open-loop gain value; the reinforcement learning module is used to read the simulation test result, and judge whether the simulation test result meets the open-loop gain value. Gain value, if the simulation test result is completely consistent with the open-loop gain value, the simulation test result and the schematic diagram are output to the output storage module, and the storage module is used to save the simulation test result and the adjusted operational amplifier schematic diagram; if the simulation test result does not meet the open-loop gain value, the reinforcement learning module adds a penalty value to the simulation test result according to the zero-order optimization algorithm, the covariance matrix adaptive evolution strategy algorithm, the Bayesian optimization algorithm and the expert prior algorithm, automatically generates new operational amplifier device parameters, and modifies the device parameters in the schematic diagram, and outputs the modified schematic diagram to the simulation test module for re-simulation testing. Repeat the above process until the simulation test result meets the open-loop gain value, and store the schematic diagram and the simulation test result.

[0030] In actual application, the automatic parameter adjustment system and method of the operational amplifier design pre-simulation process of the present application can automatically search for ideal parameters close to each device based on the manually set initial parameter values ​​and open-loop gain values ​​based on the solution capability of reinforcement learning, so that each device can find parameter values ​​that meet the open-loop gain value, have better performance, and lower power consumption in a shorter time, thereby realizing automatic and efficient design or assisting manual design; the processing optimization of the simulation test process and the adjustment of the parameters of each device are effectively integrated, thereby improving the efficiency of the operational amplifier design pre-simulation and the reliability of the optimization scheme. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] In order to more clearly illustrate the technical solution of the present application, the drawings required for use in the embodiments are briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0032] Figure 1 A schematic diagram of an automatic parameter adjustment system for a pre-simulation process of an operational amplifier design according to an embodiment of the present application;

[0033] Figure 2 A schematic diagram of an automatic parameter adjustment method for a pre-simulation process of an operational amplifier design according to an embodiment of the present application. DETAILED DESCRIPTION

[0034] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. The components of the embodiments of the present application described and shown in the drawings here can be arranged and designed in various different configurations.

[0035] In the description of this application, it should also be noted that, unless otherwise clearly specified and limited, the term "connection" should be understood in a broad sense, for example, it can be an electrical connection or a communication connection. For ordinary technicians in this field, the specific meanings of the above terms in this application can be understood according to specific circumstances.

[0036] In addition, the terms "including" and "for" and any variations thereof are intended to cover but not exclude inclusion, for example, a product or device comprising a list of components is not necessarily limited to those components explicitly listed but may include other components not explicitly listed or inherent to such products or devices.

[0037] In order to solve the problem that the operational amplifier design in the prior art requires simulation testing, manual operation consumes a lot of energy, is inefficient and has a slow design process, see Figure 1 , which is a schematic diagram of an automatic parameter adjustment system for an operational amplifier design pre-simulation process according to an embodiment of the present application; The first aspect of an embodiment of the present application provides an automatic parameter adjustment system for an operational amplifier design pre-simulation process, including: a drawing input module, a simulation test module, a reinforcement learning module and an output storage module.

[0038] In an embodiment of the present application, the drawing input module is used to draw the operational amplifier schematic diagram, and the simulation test module is used to target the open-loop gain value, read the drawn operational amplifier schematic diagram, perform simulation test and output the simulation test results; specifically, the user creates a file in the drawing input module, names it, draws the schematic diagram to be tested, edits the adjustment parameters, that is, the parameter values ​​of each component that needs to be adjusted in the schematic diagram, such as W and L of the CMOS tube, the current source and voltage source size; sets the simulation type, such as AC simulation, DC simulation and transient simulation; designs the open-loop gain value; and then performs a circuit simulation test. After the simulation is completed, the simulation test module will generate a corresponding folder in the folder to store the simulation results, which stores the schematic diagram file and circuit netlist file generated by the simulation, as well as the simulation analysis results, including all static operating points, all node voltages and currents of AC simulation, DC simulation and transient simulation analysis.

[0039] Furthermore, in some embodiments of the present application, the reinforcement learning module is used to read the simulation test results and determine whether the simulation test results meet the open-loop gain value; the open-loop gain value is a performance indicator that the operational amplifier simulation results need to meet, including open-loop gain, bandwidth, phase margin, slew rate, output swing, settling time, input impedance, output impedance, noise, power consumption, power supply rejection ratio, and common mode rejection ratio.

[0040] In an embodiment of the present application, if the simulation test results are completely consistent with the open-loop gain value, the simulation test results and the schematic diagram are output to the output storage module, and the storage module is used to save the simulation test results and the operational amplifier schematic diagram after parameter adjustment; that is, the saved simulation test results and operational amplifier schematic diagram are the final results; when optimizing the operational amplifier design, the reinforcement learning module receives the simulation test results obtained from the simulation test module.

[0041] Furthermore, in some embodiments of the present application, if the simulation test results do not meet the open-loop gain value, the reinforcement learning module automatically adjusts the schematic diagram, outputs the automatically adjusted schematic diagram to the simulation test module for re-simulation testing, and repeats the above process until the operational amplifier performance parameters obtained by the simulation test fully meet the open-loop gain value, and stores the schematic diagram and the simulation test results; in actual application, the user needs to select the reinforcement learning algorithm and the upper limit of the number of iterations for the reinforcement learning module in the automatic parameter adjustment system, manually set or select the expert experience algorithm to automatically generate the initial value of each device parameter, and set the open-loop gain value of the operational amplifier schematic design. The reinforcement learning algorithm includes a zero-order optimization algorithm, a covariance matrix adaptive evolution strategy algorithm, and a Bayesian optimization algorithm.

[0042] Further, in an embodiment of the present application, the automatic parameter adjustment is specifically to search and generate different specific parameter values ​​of each device according to the zero-order optimization algorithm, the covariance matrix adaptive evolution strategy algorithm and the Bayesian optimization algorithm; abstract the simulation test module into a callable function program, the input of the function is a set of specific parameter values ​​of each component in the operational amplifier schematic, and the output of the function is the simulation test result of the operational amplifier schematic; according to the penalty function in the expert prior algorithm, the simulation test result of the simulation test module is calculated as a specific floating-point value; the reinforcement learning module continuously calls and runs the simulation test module in the form of a function call, inputs the different specific parameter values ​​of each device generated by the zero-order optimization algorithm, the covariance matrix adaptive evolution strategy algorithm and the Bayesian optimization algorithm into the simulation test module, and submits the simulation test results to the penalty function for calculation as the penalty value; the zero-order optimization algorithm, the covariance matrix adaptive evolution strategy algorithm and the Bayesian optimization algorithm re-search and generate different specific parameter values ​​of each device according to the specific parameter values ​​of each device generated by them and their corresponding penalty values, and repeat the above process until the parameter values ​​of each device that can make the performance parameters of the operational amplifier meet the open-loop gain value are found. If the set upper limit of iterations is reached and the problem is still not found, the loop of the reinforcement learning module is forced to end.

[0043] In some embodiments of this application, the automatic parameter adjustment is specifically implemented as follows:

[0044] 1. The sampling model D is established according to the initial values ​​of each device parameter by the reinforcement learning algorithm.

[0045] 2. The reinforcement learning algorithm samples from the model D to generate multiple groups of different specific parameter values ​​for each device. For example, if the component parameters that need to be optimized for the operational amplifier are W and L of a CMOS tube in the circuit, and a Miller capacitor C, and the different groups of parameter values ​​generated are represented by subscripts, then the multiple groups of specific parameter values ​​for each device generated by the reinforcement learning algorithm are (W1, L1, C1; W2, L2, C2; W3, L3, C3; ...; Wm, Lm, Cm).

[0046] 3. The reinforcement learning module continuously runs the simulation test module in the form of function calls, inputs different groups of device parameter values, and obtains corresponding simulation test results. For example, if the simulation test of the operational amplifier is set to open-loop gain simulation, Gain is used to represent the specific open-loop gain value obtained by the simulation test result, and the subscript is used to represent the simulation test results of different groups of parameter values. After continuous simulation testing, (Gain1, Gain2, Gain3, ..., Gainm) will be obtained.

[0047] 4. According to the penalty function in the expert prior algorithm, the simulation test results of the simulation test module are calculated as penalty values, which are specific floating point numbers. For example, if the open-loop gain value of the operational amplifier is GainTarget, GainTarget and the open-loop gain values ​​of each group mentioned in step 3 are substituted into the penalty function, and the penalty function values ​​of each group (f1, f2, f3, ..., fm) can be obtained.

[0048] 5. The reinforcement learning algorithm constructs a new model D1 based on the multiple groups of specific parameter values ​​of each device and the penalty function values ​​corresponding to each group, and repeats the above four steps in a cycle.

[0049] 6. If a set of device parameter values ​​that can make the performance parameters of the operational amplifier meet the open-loop gain value is found or the set upper limit of iterations is reached and still not found, the automatic parameter adjustment ends.

[0050] Furthermore, in some embodiments of the present application, steps 1, 2, and 5 in the specific implementation method of automatic parameter adjustment are automatically completed by a reinforcement learning algorithm.

[0051] In an embodiment of the present application, if the simulation test results are completely consistent with the open-loop gain value, the simulation test results and the schematic diagram are output to the output storage module, and the storage module is used to save the simulation test results and the operational amplifier schematic diagram after parameter adjustment; that is, the saved simulation test results and operational amplifier schematic diagram are the final results; when optimizing the operational amplifier design, the reinforcement learning module receives the simulation test results obtained from the simulation test module.

[0052] Furthermore, in some embodiments of the present application, if the simulation test results do not meet the open-loop gain value, the reinforcement learning module pre-processes the schematic diagram, outputs the pre-processed schematic diagram to the simulation test module for re-simulation testing, until the operational amplifier performance parameters obtained by the simulation test completely meet the open-loop gain value, and stores the schematic diagram and the simulation test results; in actual application, the user selects the iterative algorithm and the number of iterations in the automatic parameter adjustment system, and sets the open-loop gain value of the automatic parameter adjustment, that is, the target value that each device parameter needs to reach after parameter adjustment; the automatic parameter adjustment system can continuously repeat the simulation test and process the schematic diagram simulation test results, and automatically generate new operational amplifier device parameters by the selected reinforcement learning iterative algorithm, until the operational amplifier performance parameters obtained by the simulation test results reach the open-loop gain value set by the user in the automatic parameter adjustment system or the number of cycles reaches the number of iterations; the final iteration result is output as a table for the user to choose to save.

[0053] Further, in the embodiment of the present application, the preprocessing is specifically to abstract the simulation test module into a function according to the zero-order optimization algorithm, the covariance matrix adaptive evolution strategy algorithm, the Bayesian optimization algorithm and the expert prior algorithm, the input of the function is the parameter value of each component in the operational amplifier schematic diagram, and the output of the function is calculated as a specific floating point value by the penalty function in the expert prior algorithm. The function will be solved by the zero-order optimization algorithm, the covariance matrix adaptive evolution strategy algorithm, and the Bayesian optimization algorithm to find out the parameter values ​​of each device that can make the performance parameters of the operational amplifier meet the open-loop gain value.

[0054] Specifically, in some embodiments of the present application, the reinforcement learning module preprocesses the schematic diagram. The reinforcement learning module will read the initial parameter values ​​of the parameters of each component manually set by the user when the simulation test module is first run as the input variable of the function according to the initialization parameters such as the iterative algorithm, number of iterations and search range of each device parameter selected by the user, optimize the entire schematic diagram, continuously search and try new parameter values ​​of each device parameter, perform repeated simulation tests, and realize iterative search; according to the zero-order optimization algorithm, the covariance matrix adaptive evolution strategy algorithm, the Bayesian optimization algorithm and the expert prior algorithm, the reinforcement learning module solves the function abstracted from the simulation test module. The basic process is:

[0055] Establish a sampling model D according to the initial values ​​of each device parameter;

[0056] Sample a set of solutions (such as x1, x2, x3, ...) from model D;

[0057] According to each solution xi, calculate the corresponding function value f(xi);

[0058] Construct a new model D based on the obtained solution set and its corresponding function value;

[0059] According to the new model D, the solution set and its corresponding function value are re-obtained until the convergence condition of the iterative algorithm is met;

[0060] The obtained optimal solution is output, that is, the device parameter values ​​that can make the performance parameters of the operational amplifier meet the open-loop gain value.

[0061] In the embodiment of the present application, the adopted zero-order optimization algorithm converts the constrained optimization problem into an unconstrained optimization problem by approximating the objective function or adding a penalty function to the objective function. The zero-order optimization algorithm does not use the first-order derivative information, so the zero-order optimization algorithm requires less time; the adopted covariance matrix adaptive evolution strategy algorithm simulates the biological evolution process in nature to achieve the optimization purpose. The algorithm has the characteristics of good global performance and high optimization efficiency, and provides a new way to solve the optimization problem of operational amplifier design with high computational cost; the adopted Bayesian optimization solves the extreme value problem of the function with unknown expression, constructs the acquisition function according to the result of the algorithm Gaussian process regression, solves the extreme value of the acquisition function to determine the next sampling point, and thus obtains the extreme value of the function. Its characteristics have strong sample validity, and only a small number of iterations are required to obtain a better result. The above three reinforcement learning iterative algorithms all adopt open source solutions.

[0062] In some embodiments of the present application, an expert prior algorithm is introduced to help the iterative algorithm to search more quickly and effectively for the device parameter values ​​that can make the performance parameters of the operational amplifier conform to the open-loop gain value, the initial value of the parameter adjustment variable, that is, the parameter value set by the user during the simulation test, and the appropriate initial value helps to better adjust the parameter; and the search range of each device parameter value, that is, the search interval for the target parameter value of each device parameter searched by the reinforcement learning algorithm, the appropriate range helps to achieve a balance between the time spent on parameter adjustment and the results of the parameter adjustment; the penalty function penalty value calculation formula of the performance parameters obtained from the simulation test of each parameter adjustment, the return value of the penalty function is obtained by calculating the open-loop gain value and the operational amplifier performance parameters obtained from the simulation test, and the appropriate penalty function penalty value calculation formula helps reinforcement learning to converge better.

[0063] In some embodiments of the present application, reinforcement learning is developed from theories such as animal learning and parameter perturbation adaptive control. Its basic principle is: if a behavioral strategy of the characteristics of the software and hardware system leads to a positive reward in the environment, then the tendency of the characteristics of the software and hardware system to produce this behavioral strategy in the future will be strengthened. The goal of the characteristics of the software and hardware system is to find the optimal strategy in each discrete state to maximize the expected discounted reward; reinforcement learning regards learning as a trial and evaluation process, and the characteristics of the software and hardware system select an action for the environment. After the environment receives the action, the state changes and a reinforcement signal is generated at the same time, such as an increase in the probability of reward or punishment; the selected action not only affects the immediate reinforcement value, but also affects the state of the environment at the next moment and the final reinforcement value.

[0064] It can be seen from the above technical scheme that the present application provides an automatic parameter adjustment system for the pre-simulation process of operational amplifier design, including: a drawing input module, a simulation test module, a reinforcement learning module and an output storage module; the drawing input module is used to draw the operational amplifier schematic diagram, the simulation test module is used to read the drawn operational amplifier schematic diagram with the open-loop gain value as the target, perform simulation test and output the simulation test result; the simulation test result is the operational amplifier performance parameter, and the operational amplifier performance parameter corresponds to the open-loop gain value; the reinforcement learning module is used to read the simulation test result, and judge whether the simulation test result meets the requirements. The open-loop gain value is obtained. If the simulation test result completely meets the open-loop gain value, the simulation test result and the schematic diagram are output to the output storage module, and the storage module is used to save the simulation test result and the adjusted operational amplifier schematic diagram; if the simulation test result does not meet the open-loop gain value, the reinforcement learning module adds a penalty value to the simulation test result according to the zero-order optimization algorithm, the covariance matrix adaptive evolution strategy algorithm, the Bayesian optimization algorithm and the expert prior algorithm, automatically generates new operational amplifier device parameters, and modifies the device parameters in the schematic diagram, and outputs the modified schematic diagram to the simulation test module for re-simulation testing. The above process is repeated cyclically until the simulation test result meets the open-loop gain value, and the schematic diagram and the simulation test result are stored.

[0065] In actual application, the automatic parameter adjustment system and method of the operational amplifier design pre-simulation process of the present application can automatically search for and approximate the ideal parameters of each device based on the manually set initial parameter values ​​and open-loop gain values ​​based on the solution capability of reinforcement learning, so that each device can find parameter values ​​that meet the open-loop gain value, have better performance, and lower power consumption in a shorter time, thereby realizing automatic and efficient design or assisting manual design; the processing optimization of the simulation test process and the adjustment of the parameters of each device are effectively integrated, thereby improving the efficiency of the operational amplifier design pre-simulation and the reliability of the optimization scheme.

[0066] In some embodiments of the present application, the automatic parameter adjustment system of the present application will read the initial value of each component parameter set by the user during the initial simulation test, and multiply the initial value of each component parameter by different coefficients according to the common search range of each component parameter of each parameter with different numerical values ​​as the recommended parameter adjustment variable search range of the parameter. For component parameters with initial values ​​in [0, 50), multiply by a coefficient of 0.5 as the search range, for example, for a capacitor with an initial value of 10pF, the search range is set to [7.5, 12.5]; for component parameters with initial values ​​in [50, 1000), multiply by a coefficient of 0.8 as the search range, for example, if the initial value W of the MOS tube is 100, the search range is set to [60, 140]; for component parameters with initial values ​​greater than or equal to 1000, multiply by a coefficient of 1 as the search range, for example, if the initial value W of the MOS tube is 2000, the search range is set to [1000, 3000]. At the same time, users can modify the parameter adjustment variable search range of each parameter according to their needs.

[0067] In some embodiments of the present application, the expert prior algorithm includes a penalty function, which is calculated by the open-loop gain value and various performance parameters of the current operational amplifier; when dealing with problems with multiple open-loop gain values, the same penalty function curve is usually used for each performance parameter.

[0068] Furthermore, in some embodiments of the present application, the penalty function is specifically:

[0069]

[0070]

[0071] Where: f is the return value of the penalty function, f i is one of the penalty function summation formulas, N is the number of open-loop gain values, x i are the open-loop gain values, is the performance parameter corresponding to each open-loop gain value.

[0072] In some embodiments of the present application, the reinforcement learning module will use different penalty value calculation formulas for each performance parameter based on the open-loop gain value of the automatic parameter adjustment set by the user, according to the expert prior algorithm, and according to the numerical range of each open-loop gain value, so as to avoid, during the calculation process of the penalty function return value of the reinforcement learning algorithm, when each performance parameter differs from its corresponding open-loop gain value in the same proportion, a larger penalty value calculated for a performance parameter with a larger value, and a smaller penalty value calculated for a performance parameter with a smaller value. Since the reinforcement learning algorithm tends to find component parameter values ​​that make the penalty function return value smaller, the final optimization result of the performance parameter corresponding to the open-loop gain value with a larger value may not meet the set open-loop gain value.

[0073] Specifically, in some embodiments of the present application, if the open-loop gain value set by the user is an open-loop gain of 200dB and a bandwidth of 1MHz, the reinforcement learning algorithm is run, and the program continuously searches for new parameter values ​​of each component and uses the command line to call the simulation test to obtain the value of the performance parameter. Suppose a certain set of parameter values ​​of each component makes the performance parameter value an open-loop gain of 198dB and a bandwidth of 1MHz, and the corresponding penalty function return value is 4; suppose another set of parameter values ​​of each component makes the performance parameter value an open-loop gain of 200dB and a bandwidth of 0.1MHz, and the corresponding penalty function return value is 0.81. The reinforcement learning algorithm will continuously search for new parameter values ​​of each component around the parameter values ​​of each component that make the return value of the penalty function smaller, that is, find the parameter values ​​of each component that minimize the return value of the penalty function.

[0074] In some embodiments of the present application, different penalty value calculation formulas are used for each performance parameter, so as to adjust the influence of each open-loop gain value on the calculation of the penalty function return value.

[0075] Specifically, in some embodiments of the present application, the penalty value calculation of each performance parameter is multiplied by a coefficient or the penalty function calculation formula is replaced according to the numerical value of each open-loop gain value. For example, when the difference between the open-loop gain value with the largest value and the open-loop gain value with the smallest value is greater than 40 times, the penalty value calculation formula corresponding to each open-loop gain value with the smallest value is multiplied by 1.0001; when the difference between the open-loop gain value with the largest value and the open-loop gain value with the smallest value is greater than 1000 times, the penalty value calculation formula corresponding to the open-loop gain value with the smallest value is replaced with the penalty function described herein.

[0076] Further, in some embodiments of the present application, the corresponding item of the open-loop gain value with the smallest value in the penalty function is specifically:

[0077]

[0078] Where: f i is one of the penalty function summation formulas, x i is the open-loop gain value with the minimum value in the penalty function, is the performance parameter corresponding to the open-loop gain value.

[0079] In some embodiments of the present application, the operational amplifier schematic diagram also includes: component graphic symbols, component text symbols, component netlist, component coordinates, component rotation angle, coordinates of each pin of the component, connecting lines, component position number coordinates, circuit node coordinates and text annotations.

[0080] In some embodiments of the present application, the above steps are all completed under a Linux computer system.

[0081] In order to realize the practical application of the above system, the second aspect of the embodiment of the present application also provides an automatic parameter adjustment method for the pre-simulation process of operational amplifier design, see Figure 2 , is a schematic diagram of an automatic parameter adjustment method for an operational amplifier design pre-simulation process according to an embodiment of the present application, the automatic parameter adjustment method for operational amplifier design includes: drawing an operational amplifier schematic diagram, which is used to target an open-loop gain value, reading the drawn operational amplifier schematic diagram, performing a simulation test and outputting a simulation test result; the drawing input module is used to draw an operational amplifier schematic diagram, which is used to target an open-loop gain value, reading the drawn operational amplifier schematic diagram, performing a simulation test and outputting a simulation test result; reading the simulation test result, and judging whether the simulation test result meets the open-loop gain value, if the simulation test result completely meets the open-loop gain value, outputting the simulation test result and the schematic diagram to the output storage module, the storage module is used to save the simulation test result and the operational amplifier schematic diagram after parameter adjustment; if the simulation test result does not meet the open-loop gain value, the reinforcement learning module adds a penalty value to the simulation test result according to the zero-order optimization algorithm, the covariance matrix adaptive evolution strategy algorithm, the Bayesian optimization algorithm and the expert prior algorithm, automatically generates new parameters of each device of the operational amplifier, and modifies the parameters of each device in the schematic diagram, and outputs the modified schematic diagram to the simulation test module for re-simulation testing. The above process is repeated cyclically until the simulation test result meets the open-loop gain value, and the schematic diagram and the simulation test result are stored.

[0082] A third aspect of some embodiments of the present application provides a computer device comprising:

[0083] Memory, used to store computer programs.

[0084] A processor is used to implement the steps of an automatic parameter adjustment method based on an operational amplifier design pre-simulation process as described in the second aspect of the embodiment of the present application when executing the computer program.

[0085] From the above technical scheme, it can be known that an automatic parameter adjustment system for the pre-simulation process of operational amplifier design includes: a drawing input module, a simulation test module, a reinforcement learning module and an output storage module; the drawing input module is used to draw the operational amplifier schematic diagram, the simulation test module is used to read the drawn operational amplifier schematic diagram with the open-loop gain value as the target, perform simulation test and output the simulation test result; the simulation test result is the operational amplifier performance parameter, and the operational amplifier performance parameter corresponds to the open-loop gain value; the reinforcement learning module is used to read the simulation test result, and judge whether the simulation test result meets the requirements of the open-loop gain value. If the simulation test result completely meets the open-loop gain value, the simulation test result and the schematic diagram are output to the output storage module, and the storage module is used to save the simulation test result and the adjusted operational amplifier schematic diagram; if the simulation test result does not meet the open-loop gain value, the reinforcement learning module adds a penalty value to the simulation test result according to the zero-order optimization algorithm, the covariance matrix adaptive evolution strategy algorithm, the Bayesian optimization algorithm and the expert prior algorithm, automatically generates new operational amplifier device parameters, and modifies the device parameters in the schematic diagram, and outputs the modified schematic diagram to the simulation test module for re-simulation testing. The above process is repeated until the simulation test result meets the open-loop gain value, and the schematic diagram and the simulation test result are stored.

[0086] In actual application, the automatic parameter adjustment system and method of the operational amplifier design pre-simulation process of the present application can automatically search for and approximate the ideal parameters of each device based on the manually set initial parameter values ​​and open-loop gain values ​​based on the solution capability of reinforcement learning, so that each device can find parameter values ​​that meet the open-loop gain value, have better performance, and lower power consumption in a shorter time, thereby realizing automatic and efficient design or assisting manual design; the processing optimization of the simulation test process and the adjustment of the parameters of each device are effectively integrated, thereby improving the efficiency of the operational amplifier design pre-simulation and the reliability of the optimization scheme.

[0087] The present application is described in detail above in conjunction with specific implementation methods and exemplary examples, but these descriptions cannot be understood as limiting the present application. Those skilled in the art understand that, without departing from the spirit and scope of the present application, various equivalent substitutions, modifications or improvements can be made to the technical solutions and implementation methods of the present application, all of which fall within the scope of the present application.

Claims

1. An automatic parameter adjustment system for the pre-simulation process of operational amplifier design. It is characterized in that include: Draw input module, simulation test module, reinforcement learning module and output storage module; The drawing input module is used to draw the operational amplifier schematic diagram, and the simulation test module is used to read the drawn operational amplifier schematic diagram with a preset open-loop gain value as a target, perform simulation test and output simulation test results; the simulation test results are operational amplifier performance parameters, and the operational amplifier performance parameters correspond to the open-loop gain value; The reinforcement learning module is used to read the simulation test result and determine whether the simulation test result meets the open-loop gain value. If the simulation test result completely complies with the open-loop gain value, the simulation test result and the schematic diagram are output to the output storage module, and the storage module is used to store the simulation test result and the operational amplifier schematic diagram after parameter adjustment; If the simulation test result does not meet the open-loop gain value, the reinforcement learning module adds a penalty value to the simulation test result according to the zero-order optimization algorithm, the covariance matrix adaptive evolution strategy algorithm, the Bayesian optimization algorithm and the expert prior algorithm, automatically generates new parameters of each device of the operational amplifier, modifies the parameters of each device in the schematic diagram, outputs the modified schematic diagram to the simulation test module for re-simulation test, and repeats the above process in a cycle until the simulation test result meets the open-loop gain value, stores the schematic diagram and the simulation test result, and the expert prior algorithm includes a penalty function, which is obtained by calculating the open-loop gain value and the operational amplifier performance parameters obtained by the simulation test; The penalty function is specifically: Where: f is the return value of the penalty function, f i is one of the penalty function summation formulas, N is the number of open-loop gain values, x i are the open-loop gain values, is the performance parameter corresponding to each open-loop gain value.

2. The automatic parameter adjustment system for the pre-simulation process of an operational amplifier design according to claim 1, It is characterized in that The penalty function is: Where: f i is one of the penalty function summation formulas, x i is the open-loop gain value with the minimum value in the penalty function, is the performance parameter corresponding to the open-loop gain value.

3. The automatic parameter adjustment system for the pre-simulation process of an operational amplifier design according to claim 1, It is characterized in that The operational amplifier schematic diagram also includes: component graphic symbols, component text symbols, component netlist, component coordinates, component rotation angles, coordinates of each pin of the component, connecting lines, component bit number coordinates, circuit node coordinates and text annotations.

4. The automatic parameter adjustment system for the pre-simulation process of an operational amplifier design as claimed in claim 1, Features: The above steps are all completed under the Linux computer system.

5. An automatic parameter adjustment method for the pre-simulation process of operational amplifier design. It is characterized in that An automatic parameter adjustment system for an operational amplifier design pre-simulation process according to any one of claims 1 to 4, wherein the automatic parameter adjustment method for the operational amplifier design comprises: Draw an operational amplifier schematic diagram, use the open-loop gain value as a target, read the drawn operational amplifier schematic diagram, perform simulation tests and output simulation test results; Reading the simulation test result, and determining whether the simulation test result meets the open-loop gain value, If the simulation test result completely complies with the open-loop gain value, the simulation test result and the schematic diagram are output to the output storage module, and the storage module is used to store the simulation test result and the operational amplifier schematic diagram after parameter adjustment; If the simulation test result does not meet the open-loop gain value, the reinforcement learning module adds a penalty value to the simulation test result according to the zero-order optimization algorithm, the covariance matrix adaptive evolution strategy algorithm, the Bayesian optimization algorithm and the expert prior algorithm, automatically generates new parameters of each device of the operational amplifier, and modifies the parameters of each device in the schematic diagram. The modified schematic diagram is output to the simulation test module for re-simulation testing, and the above process is repeated in a cycle until the simulation test result meets the open-loop gain value, and the schematic diagram and the simulation test result are stored.

6. A computer device, It is characterized in that include: Memory for storing computer programs; A processor is used to implement the steps of the automatic parameter adjustment method based on an operational amplifier design pre-simulation process as claimed in claim 5 when executing the computer program.

Images