Switching power supply aided design method and system

Abstract

The invention provides a switching power supply aided design method and system, and the method comprises the steps: carrying out simulation and physical experiment tests according to influence parameters and an experiment circuit, and obtaining the data of each influence parameter under each test; performing linear regression on each influence parameter under each test to obtain a regression function about the output voltage ripple and other influence parameters; according to the causal inference result, obtaining all influence parameters with causal relationships and a causal effect value corresponding to each causal relationship; defining each influence parameter as an entity, summarizing entity relationships among all entities, and associating corresponding regression coefficients or causal effect values according to the entity relationships to obtain a power supply knowledge graph; and obtaining power supply design requirements, searching parameters to be debugged from the power supply knowledge graph according to the power supply design requirements, and optimizing the power supply design scheme according to the parameters to be debugged. The design period of the switching power supply can be greatly shortened, and the intelligent process of power supply design is promoted.

Description

Technical Field

[0001] The present invention relates to the technical field of power supply auxiliary design, and in particular to a switching power supply auxiliary design method and system. Background Art

[0002] With the demand for miniaturization and high efficiency of electronic devices, low power consumption design and high energy efficiency conversion of switching power supplies are becoming increasingly important. However, traditional design methods make it difficult to fully consider the complex impact of multiple factors on power supply performance, and the design process is complicated and time-consuming. Summary of the invention

[0003] The purpose of the present invention is to provide a switching power supply auxiliary design method and system, aiming to solve the problem that traditional technology is difficult to fully consider the complex influence of multiple factors on power supply performance, resulting in a complex and time-consuming design process.

[0004] In a first aspect, the present invention provides a switching power supply auxiliary design method, the method comprising: Build an experimental circuit, select multiple influencing parameters that affect power supply performance, and perform simulation and physical experimental tests based on the influencing parameters and the experimental circuit to obtain data of each influencing parameter under each test; Performing linear regression on each influencing parameter under each test to obtain a regression function of the output voltage ripple and other influencing parameters, and determining whether the regression function meets a preset fitting condition; If the regression function meets the preset fitting condition, the regression coefficients of the output voltage ripple and other influencing parameters are obtained according to the regression function, and causal discovery is performed on any two influencing parameters based on the PC algorithm, and all influencing parameters with causal relationships are obtained according to the causal discovery results, and then the causal effect values ​​corresponding to the influencing parameters and each causal relationship are obtained through the structural equation model; Define each influencing parameter as an entity, summarize the entity relationships between all entities, and associate the corresponding regression coefficients or causal effect values ​​according to the entity relationships to obtain the power supply knowledge graph; The power supply design requirements are obtained, and the parameters to be debugged are searched from the power supply knowledge graph according to the power supply design requirements, and the power supply design scheme is optimized according to the parameters to be debugged.

[0005] In summary, according to the above-mentioned switching power supply auxiliary design method, the root cause of the problem can be quickly located through the knowledge graph query method, and the design auxiliary method based on causal inference and knowledge graph can systematically reveal the causal relationship and provide a more accurate optimization solution, thereby greatly improving the efficiency, accuracy and intelligence level of the design. Especially in complex power supply design, the advantages of the embodiments of the present invention are particularly prominent, which not only improves the quality of design decisions, but also greatly shortens the design cycle and promotes the intelligent process of power supply design.

[0006] Furthermore, the influencing parameters include inductor core material, output capacitance, switching frequency, input voltage, output voltage, output voltage ripple, inductor current ripple, maximum limiting current, and control voltage; The physical experiment test includes a first test experiment and a second test experiment, and the steps of the first test experiment include: Step 1.1, select 220μF output capacitor, select 15μH powder core inductor as inductor core material, set switching frequency to 500kHz, turn on oscilloscope, turn on power supply, set voltage to 12V, and set current limit to 1.5A, turn on electronic load, set constant current mode, and set current to 1.5A; Step 2.1, turn on the power "OUT ON" button, and the waveform of the output voltage Vout displayed on the oscilloscope is a flat horizontal line of 3.3V; Step 3.1. Turn on the "LOAD ON" button of the electronic load. At this time, a triangle waveform with an average value of 1.5A is displayed on the inductor current channel of the oscilloscope, and a flat horizontal line with a level of 3.3V is displayed on the output voltage channel Vout. Step 4.1. Read the output voltage Vout, measure the output voltage ripple and inductor current ripple, and record these values. Then change the input voltage Vin and load current values ​​and repeat this step. Step 5.1. Turn off the "LOAD ON" button of the electronic load and the "OUT ON" button of the DC power supply, then connect a 15μH ferrite core inductor and repeat steps 2.1 to 4.1. Step 6.1, set the switching frequency to 500kHz, and repeat steps 2.1 to 5.1; Step 7.1. At the end of the measurement, turn off the “LOAD ON” button of the electronic load and the “OUT ON” button of the DC power supply, and then turn off all instruments.

[0007] Furthermore, the steps of the second test experiment include: Step 1.2, select an output capacitor of 220μF, select a powder core inductor with a core material X of 15μH, set the switching frequency to 250kHz, turn on the oscilloscope, turn on the power supply, set the voltage to 12V, and set the current limit to 1.5A, turn on the electronic load, set the constant current mode, and set the current to 1A; Step 2.2, turn on the power "OUT ON" button, and the output voltage waveform displayed on the oscilloscope is a flat horizontal line of 3.3V; Step 3.2, turn on the "LOAD ON" button of the electronic load. At this time, a triangle waveform with an average value of 1A is displayed on the inductor current channel of the oscilloscope, a flat horizontal line with a level between 500mV and 1V and very small ripple is seen on the control voltage channel, and a flat horizontal line with a level of 3.3V is seen on the output voltage channel; Step 4.2, slowly increase the current of the electronic load, and detect the maximum current limit value when the output voltage is stabilized at 3.3V. When the output voltage is no longer stable, turn off the "LOAD ON" button of the electronic load and the "OUT ON" button of the power supply, then record the maximum current limit and the corresponding control voltage value, change the input voltage value, and repeat this step; Step 5.2, select a 15μH ferrite core inductor or a 15μH magnetic powder core inductor, a 10μF output capacitor or a 220μF output capacitor, a 250kHz switching frequency or a 500kHz switching frequency, and repeat steps 2.2 to 4.2 by cross-combining the three; Step 6.2: At the end of the measurement, turn off the “LOAD ON” button of the electronic load and the “OUT ON” button of the DC power supply, and then turn off all instruments.

[0008] Furthermore, the steps of performing linear regression on each influencing parameter under each test to obtain a regression function of the output voltage ripple and other influencing parameters, and judging whether the regression function satisfies a preset fitting condition include: Linear regression is performed according to the following formula: ; in, Indicates the inductor core material, Indicates the input voltage, Indicates the maximum limiting current, represents the control voltage, represents the switching frequency, Indicates the output voltage, represents the output voltage ripple, represents the inductor current ripple, represents the output capacitance, represents the intercept term, , , , , , , , are regression coefficients, represents the error term; The preset fitting conditions include whether the multivariate determination coefficient after the regression function is adjusted is greater than a first preset threshold, and whether the p-value of the regression function is greater than a second preset threshold; The adjusted multivariate determination coefficient was calculated according to the following formula: ; in, represents the adjusted multivariate determination coefficient, represents the original multivariate determination coefficient, represents the experimental data capacity, Indicates the number of independent variables; The original multivariate determination coefficient was calculated according to the following formula: ; in, represents the residual sum of squares, represents the total sum of squares, represents the fitting value of the qth experimental data, represents the average value of the experimental data, represents the qth experimental data, and q represents the number of experimental data.

[0009] Furthermore, the step of performing causal discovery on any two influencing parameters based on the PC algorithm and obtaining all influencing parameters with causal relationships according to the causal discovery results includes: Determine each pair of influencing parameters and Whether it is independent under a given set of variables Y: ; If independent, delete and The edges between them are identified, and the directions of the edges are determined by identifying the V-shaped structure and applying the transitivity rule, thus obtaining a directed acyclic graph used to represent the causal relationship between the influencing parameters.

[0010] Furthermore, the step of obtaining the causal effect value corresponding to the influencing parameter and each causal relationship through the structural equation model includes: The causal effect value is calculated according to the following formula: ; in, , Respectively represent the i-th and j-th influencing parameters, represents the error term, Influencing parameters Influencing parameters The causal effect value of Influencing parameters and influencing parameters The covariance between Influencing parameters The variance of .

[0011] Furthermore, the steps of obtaining power supply design requirements, searching for parameters to be debugged from the power supply knowledge graph according to the power supply design requirements, and optimizing the power supply design solution according to the parameters to be debugged include: Acquire at least one power supply design requirement, each of the power supply design requirements including a target impact parameter and a value corresponding to the target impact parameter; The parameters to be debugged that are related to each of the target influencing parameters are searched from the power supply knowledge graph, and the power supply design results are optimized according to the parameters to be debugged to obtain the optimal design solution.

[0012] In a second aspect, the present invention provides a switching power supply auxiliary design system, the system comprising: A simulation and physical experiment platform building module is used to build an experimental circuit, select multiple influencing parameters that affect the performance of the power supply, and perform simulation and physical experiment tests based on the influencing parameters and the experimental circuit to obtain data of each influencing parameter under each test; A linear regression module is used to perform linear regression on each influencing parameter under each test, obtain a regression function of the output voltage ripple and other influencing parameters, and determine whether the regression function meets a preset fitting condition; A causal inference module, for obtaining the regression coefficients of the output voltage ripple and other influencing parameters according to the regression function if the regression function satisfies the preset fitting condition, and performing causal discovery on any two influencing parameters based on the PC algorithm, and obtaining all influencing parameters with causal relationships according to the causal discovery results, and then obtaining the causal effect values ​​corresponding to the influencing parameters and each causal relationship through the structural equation model; The knowledge graph construction module is used to define each influencing parameter as an entity, summarize the entity relationships between all entities, and associate the corresponding regression coefficients or causal effect values ​​according to the entity relationships to obtain the power knowledge graph; The auxiliary design module is used to obtain power supply design requirements, search for parameters to be debugged from the power supply knowledge graph according to the power supply design requirements, and optimize the power supply design solution according to the parameters to be debugged.

[0013] In a third aspect, the present invention provides a storage medium, wherein the storage medium stores one or more programs, which, when executed by a processor, implement the above-mentioned switching power supply auxiliary design method.

[0014] In a fourth aspect, the present invention provides an electronic device, the electronic device comprising a memory and a processor, wherein: The memory is used to store computer programs; When the processor is used to execute the computer program stored in the memory, the above-mentioned switching power supply auxiliary design method is implemented. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 A flow chart of a switching power supply auxiliary design method proposed in one embodiment of the present invention; Figure 2 A simplified schematic diagram of a buck regulator according to an embodiment of the present invention; Figure 3 A schematic diagram of a cause-effect relationship that has been clarified based on prior knowledge according to an example of an embodiment of the present invention; Figure 4 A schematic diagram of the cause-effect relationship among five variables, namely, inductor core material, output capacitance, output voltage, switching frequency, and maximum limiting current, according to an example of an embodiment of the present invention; Figure 5 A general diagram of cause and effect relationships for an example of an embodiment of the present invention; Figure 6 A cause-and-effect diagram of an inductor core material and a power supply performance index according to an embodiment of the present invention; Figure 7 A schematic diagram of the structure of a power supply knowledge graph containing causal relationships according to an example of an embodiment of the present invention; Figure 8 A schematic diagram of factors affecting output voltage ripple according to an embodiment of the present invention; Fig. 9 (a) is a schematic diagram showing the influence of the switching frequency on other performance indicators according to an embodiment of the present invention. Fig. 9 (b) is a schematic diagram showing the influence of the inductor core material on other performance indicators according to an embodiment of the present invention. Fig. 9 (c) is a schematic diagram showing the influence of the control voltage on other performance indicators according to an embodiment of the present invention; Fig.10 A schematic diagram of the structure of a switching power supply auxiliary design system proposed in one embodiment of the present invention.

[0016] The following specific implementation manner will further illustrate the present invention in conjunction with the above-mentioned drawings. DETAILED DESCRIPTION

[0017] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. Unless otherwise defined, the technical terms or scientific terms used herein should be understood by people with general skills in the field to which the present invention belongs. "Including" and similar words used in this article mean that the elements or objects appearing before the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects.

[0018] like Figure 1 As shown, an embodiment of the present invention provides a switching power supply auxiliary design method, the method comprising steps S101 to S105, wherein: Step S101: constructing an experimental circuit, selecting multiple influencing parameters that affect power supply performance, and performing simulation and physical experimental tests based on the influencing parameters and the experimental circuit to obtain data of each influencing parameter under each test; It should be noted that the Buck circuit is one of the most common switching power supply topologies and is widely used in various electronic devices. The study of the influence and causal relationship in the Buck circuit has broad practical significance and can provide a reference for the design of other types of switching power supplies. TPS54160 is a buck-type integrated regulator chip with an internal NMOS switch. The input voltage range is 3.5V to 60V, the maximum output current is 1.5A, and the output voltage is stable at 3.3V. The switching frequency range is 100kHz to 2MHz. It is used to study the influence of different inductor core material types and operating conditions under core saturation on the inductor current ripple. The simplified schematic diagram of the buck regulator is shown in the figure. Figure 2 shown.

[0019] In addition, the simulation circuit diagram was built using the Texas Instruments (TI) power supply online experiment platform WEBENCH for simulation experiments, using two common inductor core materials: ferrite and magnetic powder core. The experimental results show that under the same nominal inductance conditions and saturation conditions, the current ripple generated by the magnetic powder core inductor at heavy load current is smaller than that generated by the ferrite inductor. When the load current increases, the relative increase in the ripple amplitude of the ferrite inductor is higher than that of the magnetic powder core inductor. In order to verify the simulation results, a physical experiment is then carried out.

[0020] Using a regular circuit board, perform the following two tests: In order to obtain the data of seven variables, namely, inductor core material, output capacitance, switching frequency, input voltage, output voltage, output voltage ripple, and inductor current ripple, the first test experiment was carried out. The steps are as follows: Step 1.1, select 220μF output capacitor, select 15μH powder core inductor as inductor core material, set switching frequency to 500kHz, turn on oscilloscope, turn on power supply, set voltage to 12V, and set current limit to 1.5A, turn on electronic load, set constant current mode, and set current to 1.5A; Step 2.1, turn on the power "OUT ON" button, and the waveform of the output voltage Vout displayed on the oscilloscope is a flat horizontal line of 3.3V; Step 3.1. Turn on the "LOAD ON" button of the electronic load. At this time, a triangle waveform with an average value of 1.5A is displayed on the inductor current channel of the oscilloscope, and a flat horizontal line with a level of 3.3V is displayed on the output voltage channel Vout. Step 4.1. Read the output voltage Vout, measure the output voltage ripple and inductor current ripple, and record these values. Then change the input voltage Vin and load current values ​​and repeat this step. Step 5.1. Turn off the "LOAD ON" button of the electronic load and the "OUT ON" button of the DC power supply, then connect a 15μH ferrite core inductor and repeat steps 2.1 to 4.1. Step 6.1, set the switching frequency to 500kHz, and repeat steps 2.1 to 5.1; Step 7.1. At the end of the measurement, turn off the “LOAD ON” button of the electronic load and the “OUT ON” button of the DC power supply, and then turn off all instruments.

[0021] In order to obtain the data of the six variables of inductor core material, output capacitance, switching frequency, input voltage, maximum limiting current, and control voltage, the second test experiment was carried out. The steps are as follows: Step 1.2, select an output capacitor of 220μF, select a powder core inductor with a core material X of 15μH, set the switching frequency to 250kHz, turn on the oscilloscope, turn on the power supply, set the voltage to 12V, and set the current limit to 1.5A, turn on the electronic load, set the constant current mode, and set the current to 1A; Step 2.2, turn on the power "OUT ON" button, and the output voltage waveform displayed on the oscilloscope is a flat horizontal line of 3.3V; Step 3.2, turn on the "LOAD ON" button of the electronic load. At this time, a triangle waveform with an average value of 1A is displayed on the inductor current channel of the oscilloscope, a flat horizontal line with a level between 500mV and 1V and very small ripple is seen on the control voltage channel, and a flat horizontal line with a level of 3.3V is seen on the output voltage channel; Step 4.2, slowly increase the current of the electronic load, and detect the maximum current limit value when the output voltage is stabilized at 3.3V. When the output voltage is no longer stable, turn off the "LOAD ON" button of the electronic load and the "OUT ON" button of the power supply, then record the maximum current limit and the corresponding control voltage value, change the input voltage value, and repeat this step; Step 5.2, select a 15μH ferrite core inductor or a 15μH magnetic powder core inductor, a 10μF output capacitor or a 220μF output capacitor, a 250kHz switching frequency or a 500kHz switching frequency, and repeat steps 2.2 to 4.2 by cross-combining the three; Step 6.2: At the end of the measurement, turn off the “LOAD ON” button of the electronic load and the “OUT ON” button of the DC power supply, and then turn off all instruments.

[0022] Comprehensive tests 1 and 2 yield a total of nine variables: inductor core material, output capacitance, switching frequency, input voltage, maximum current limit, control voltage, output voltage, output voltage ripple, and inductor current ripple.

[0023] In the physical experiment, the experimental phenomenon is consistent with the simulation results. The ferrite inductor is more likely to saturate when the current increases, resulting in an increase in current ripple. The magnetic powder core inductor shows a lower degree of saturation under the same conditions, so the current ripple is relatively small. The type of core material has a significant impact on the performance of the power supply, especially the inductor current ripple and output voltage ripple. Therefore, the reasonable selection of inductor core material is of great significance for optimizing the design of switching power supplies and improving their performance. A total of 200 sets of experimental data were collected through comprehensive simulation and physical experiments. Some of the data are shown in Table 1: Table 1 ; in, Indicates the inductor core material, is a binary variable, "1" represents ferrite core material and "0" represents magnetic powder core material. Indicates the input voltage, Indicates the maximum limiting current, represents the control voltage, represents the switching frequency, Indicates the output voltage, represents the output voltage ripple, represents the inductor current ripple, Represents the output capacitance.

[0024] Step S102: performing linear regression on each influencing parameter under each test to obtain a regression function of the output voltage ripple and other influencing parameters, and determining whether the regression function meets a preset fitting condition; It should be noted that the linear regression is performed according to the following formula: ; in, Indicates the inductor core material, Indicates the input voltage, Indicates the maximum limiting current, represents the control voltage, represents the switching frequency, Indicates the output voltage, represents the output voltage ripple, represents the inductor current ripple, represents the output capacitance, represents the intercept term, , , , , , , , are regression coefficients, represents the error term; The preset fitting conditions include whether the multivariate determination coefficient after the regression function is adjusted is greater than a first preset threshold, and whether the p-value of the regression function is greater than a second preset threshold; The adjusted multivariate determination coefficient was calculated according to the following formula: ; in, represents the adjusted multivariate determination coefficient, represents the original multivariate determination coefficient, represents the experimental data capacity, Indicates the number of independent variables; The original multivariate determination coefficient was calculated according to the following formula: ; in, represents the residual sum of squares, represents the total sum of squares, represents the fitting value of the qth experimental data, represents the average value of the experimental data, represents the qth experimental data, and q represents the number of experimental data.

[0025] The adjusted coefficient of determination is an improvement on the multivariate coefficient of determination. The adjusted coefficient of determination can more accurately reflect the fit of the model when introducing new independent variables, avoiding the overfitting of the model due to adding too many meaningless independent variables. For example, as shown in Tables 2 and 3 below, the adjusted coefficient of determination, p-value and regression coefficient of each influencing parameter are as follows: Table 2 ; Table 3 ; In Table 3, the regression coefficient between the inductor core material and the output voltage ripple is 0.6828, which means that the inductor core material has a greater impact on the output voltage ripple, while the regression coefficient between the maximum limiting current and the output voltage ripple is 0.1963, which means that the maximum limiting current has a weaker impact on the output voltage ripple. Next, we will analyze the causal relationship of the variables.

[0026] Step S103: if the regression function meets the preset fitting condition, the regression coefficients of the output voltage ripple and other influencing parameters are obtained according to the regression function, and causal discovery is performed on any two influencing parameters based on the PC algorithm, and all influencing parameters with causal relationships are obtained according to the causal discovery results, and then the causal effect values ​​corresponding to the influencing parameters and each causal relationship are obtained through the structural equation model; It should be noted that if the adjusted multivariate determination coefficient of the regression function is greater than the first preset threshold and the p-value of the regression function is greater than the second preset threshold, it means that the regression function meets the preset fitting conditions, and causal inference is performed at this time. Otherwise, it means that the preset fitting conditions are not met, that is, the obtained regression function is not good. At this time, it is necessary to reselect the fitting function until the preset fitting conditions are met.

[0027] Among the nine variables obtained in Table 1, the causal relationships that have been clarified based on prior knowledge are as follows: Figure 3 As shown, the causal relationship between input voltage → inductor current ripple, switching frequency → inductor current ripple, and inductor core material → inductor current ripple can be derived from the following formula: ; Where D is the duty cycle.

[0028] The causal relationship between switching frequency → output voltage ripple, inductor core material → output voltage ripple, and switching frequency → output voltage can be derived from the following formula: ; Where ESR is the equivalent series resistance of the output capacitor.

[0029] The characteristics of the inductor core material (such as magnetic permeability, saturation flux density, loss, etc.) directly affect the maximum limiting current of the inductor. The relationship between the saturation flux density B and the current I in the core material is: ; Among them, μ is the magnetic permeability of the core, N is the number of turns of the coil, I is the current, and l is the length of the magnetic path. That is, the cause and effect relationship of inductor core material → maximum limiting current can be derived from the above formula.

[0030] Other causal relationships are derived: Control voltage → output voltage ripple: The control voltage maintains the stability of the output voltage by adjusting the switching state of the switching device. The response speed of the control voltage directly affects the size of the output voltage ripple. If the feedback control system is well designed, the output voltage ripple can be effectively reduced. On the contrary, if the control voltage responds slowly or the feedback loop is unstable, it may cause larger ripples.

[0031] Control voltage → maximum current limit: In the peak current control mode, the control voltage directly controls the current peak of the switch tube. The current peak determines the maximum value of the inductor current, which is directly related to the maximum output current of the power supply.

[0032] Output capacitor → control voltage: The size of the capacitor directly affects the response speed of the power supply output voltage (control voltage) and its ability to withstand load changes. When the load current changes suddenly, the output capacitor helps the power supply stabilize the output voltage.

[0033] The causal relationship between the inductor core material and the output capacitance and output voltage, and the causal relationship between the switching frequency and the maximum limiting current are not yet clear. Based on the collected experimental data, the PC algorithm is used to perform causal discovery on these five variables. The PC algorithm is good at processing high-dimensional data and can accurately identify direct causal relationships between many variables. At the same time, it provides a graphical display of the relationship between variables, which is suitable for causal discovery of multiple variables in this study. The core of the PC algorithm is to use conditional independence tests to gradually construct an undirected graph G of the variable set Y. In the process of constructing the undirected graph, the PC algorithm gradually tests each pair of influencing parameters. and Whether it is independent under a given set of variables Y. If the conditions are met: , then delete and For example, the inductor core material and the input voltage are independent in the nine variable sets, that is, the edge between the inductor core material and the input voltage is deleted. Based on conditional independence, the direction of the edge is further determined by identifying the V-shaped structure and applying the transitivity rule, and finally a directed acyclic graph (DAG) is obtained.

[0034] Finally, the causal relationship between the five influencing parameters, namely, inductor core material, output capacitance, output voltage, switching frequency, and maximum limiting current, is obtained as follows: Figure 4 As shown in the figure, the black dotted arrows represent the causal relationship discovered based on the PC algorithm, and the black solid arrows represent the causal relationship known based on prior knowledge mentioned above. Figure 3 and Figure 4 The causal relationship diagram is as follows Figure 5 shown.

[0035] In addition, in some embodiments, in order to obtain the causal effect value, the sem() function in the lavaan package is introduced to fit the model and estimate the causal effect. For example, in a specific structural equation model, the maximum likelihood estimation is used to estimate the path coefficient: ; in, , Respectively represent the i-th and j-th influencing parameters, represents the error term, Influencing parameters Influencing parameters The causal effect value of Influencing parameters and influencing parameters The covariance between Influencing parameters Through the sem() function, the path coefficient can be automatically extracted after the model is fitted, and the causal effect diagram of the inductor core material and power supply performance is finally obtained. Figure 6 shown.

[0036] Step S104: define each influencing parameter as an entity, summarize the entity relationships between all entities, and associate the corresponding regression coefficients or causal effect values ​​according to the entity relationships to obtain a power knowledge graph; In this step, the causal graph is combined with the knowledge graph, so that the knowledge graph is no longer just a model of correlation, but also contains the more core causal relationships in power supply design, and finally a more comprehensive power supply knowledge graph with causal relationships is obtained. The knowledge graph modeling tool used in this study is the Neo4j graph database. The main function of Neo4j is to provide underlying storage and query capabilities. It supports the Cypher query language, which can help users easily retrieve and analyze information in the knowledge graph.

[0037] Taking the experiment of the influence of inductor core material on power supply performance as an example, a power supply knowledge graph with causal relationship is constructed. First, TI's TI-PMLK experimental instruction manual, CNKI database, literature materials, etc. are used as declarative knowledge sources, and the knowledge is extracted and integrated. Then, combined with the structural equation model, the causal graph and the triple table are matched one by one. For example, control voltage → output voltage ripple, the corresponding triple table is: entity 1: control voltage, relationship: cause and effect, entity 2: output voltage ripple, and other causal relationships are similar. A total of 54 nodes and 4 types of relationships (a total of 106 relationships) are extracted and converted to form a csv of "node-relationship-node" triples, and then the regression coefficient and causal effect data are added to the csv table. Part of the data is shown in Table 4.

[0038] Table 4 ; Finally, a power knowledge graph with causal relationships is obtained in Neo4j. The power knowledge graph with causal relationships (partial) is as follows: Figure 7 shown.

[0039] Step S105: Obtain power supply design requirements, and search for parameters to be debugged from the power supply knowledge graph according to the power supply design requirements, and optimize the power supply design solution according to the parameters to be debugged.

[0040] In this step, it is first necessary to obtain at least one power supply design requirement, each of which includes a target influencing parameter and a numerical value corresponding to the target influencing parameter; then, the parameters to be debugged that are related to each of the target influencing parameters are searched from the power supply knowledge graph, and the power supply design results are optimized according to the parameters to be debugged to obtain the optimal design solution.

[0041] For example, the Buck circuit is designed using the TI power management chip TPS54160. The power designer completes the following design tasks. The specific design and debugging tasks (design requirements) are shown in Table 5: Table 5 ; Initial measurements show that the inductor core material is ferrite material, and the actual output voltage ripple is 51mV. The design requirements are shown in Table 5. Taking the requirement "output voltage ripple ≤ 45mV" in Table 5 as an example, facing this problem, the power designer first uses the Neo4j knowledge graph to query which factors may affect the output voltage ripple. The designer can set the search for "cause and effect" relationships and obtain the factors that affect the output voltage ripple by querying the knowledge graph. Figure 8 As shown, by viewing Figure 8It can be seen that the factors that affect the output voltage ripple include the switching frequency, the inductor core material, and the control voltage. Therefore, we should first consider adjusting these three factors to reduce the output voltage ripple. Due to the complexity of the power supply, these three factors will also affect other power supply performance, resulting in other indicators being unqualified during adjustment. Therefore, the designer also needs to inquire about which other performance indicators these three factors will affect. By querying, we can get the impact of these factors on other performance indicators. Fig. 9 (a) to Fig. 9 (c). Fig. 9 (a) to Fig. 9 (c) It can be seen that there are many factors that affect the switching frequency and the inductor core material. If the switching frequency is adjusted or the inductor core material is replaced, other performance indicators may fail to meet the requirements. The control voltage only affects the maximum limiting current. Therefore, while ensuring that the maximum limiting current is qualified, the control voltage can be adjusted to reduce the output voltage ripple. Finally, the control voltage was increased to 0.97V, and the output voltage ripple was reduced to 42mV, indicating that the goal was achieved. The same method was used for other tasks. The final optimization results are shown in Table 6, that is, the task was completed: Table 6 ; In Table 5, the design and debugging tasks involve many complex influencing factors. This embodiment shows that the root cause of the problem can be quickly located through the knowledge graph query method. The design auxiliary method based on causal inference and knowledge graph can systematically reveal the causal relationship and provide a more accurate optimization solution, thereby greatly improving the efficiency, accuracy and intelligence level of the design. Especially in complex power supply design, the advantages of this research method are particularly prominent, which not only improves the quality of design decisions, but also greatly shortens the design cycle and promotes the intelligent process of power supply design.

[0042] In summary, by using knowledge graphs for power supply design, the sharing and reusability of knowledge in the power supply field is improved compared to traditional power supply knowledge acquisition; and causal inference technology helps to realize power supply design from a causal perspective, which is different from previous research methods. It crosses from the field of natural science to the field of data science and solves the power supply design ideas from the root; in addition, hidden causal relationships are discovered through causal discovery methods, which enhances the scientificity and accuracy of power supply design; and a method of combining causal inference with knowledge graphs is proposed, which expands knowledge graphs from correlation analysis to causal relationship modeling, providing new ideas for complex system design.

[0043] like Fig.10 An embodiment of the present invention further provides a switching power supply auxiliary design system, the system comprising: The simulation and physical experiment platform building module 10 is used to build an experimental circuit, select multiple influencing parameters that affect the power supply performance, and perform simulation and physical experiment tests according to the influencing parameters and the experimental circuit to obtain data of each influencing parameter under each test; The linear regression module 20 is used to perform linear regression on each influencing parameter under each test, obtain a regression function of the output voltage ripple and other influencing parameters, and determine whether the regression function meets a preset fitting condition; A causal inference module 30 is used to obtain the regression coefficients of the output voltage ripple and other influencing parameters according to the regression function if the regression function meets the preset fitting conditions, and to perform causal discovery on any two influencing parameters based on the PC algorithm, and to obtain all influencing parameters with causal relationships according to the causal discovery results, and then to obtain the causal effect values ​​corresponding to the influencing parameters and each causal relationship through the structural equation model; The knowledge graph construction module 40 is used to define each influencing parameter as an entity, summarize the entity relationships between all entities, and associate the corresponding regression coefficients or causal effect values ​​according to the entity relationships to obtain a power knowledge graph; The auxiliary design module 50 is used to obtain power supply design requirements, search for parameters to be debugged from the power supply knowledge graph according to the power supply design requirements, and optimize the power supply design solution according to the parameters to be debugged.

[0044] Another aspect of the present invention further provides an electronic device, comprising a memory and a processor, wherein the memory is used to store a computer program, and the processor is used to execute the computer program stored in the memory to implement the above-mentioned switching power supply auxiliary design method.

[0045] Those skilled in the art will appreciate that the logic and / or steps represented in the flowchart or otherwise described herein, for example, may be considered as an ordered list of executable instructions for implementing logical functions, and may be specifically implemented in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For purposes of this specification, "computer-readable medium" may be any device that can contain storage, communication, propagation or transmission of a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.

[0046] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or, if necessary, processing in another suitable manner, and then stored in a computer memory.

[0047] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or a combination thereof: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0048] Although the embodiments of the present invention are described in detail above, it is obvious to those skilled in the art that various modifications and variations can be made to these embodiments. However, it should be understood that such modifications and variations are within the scope and spirit of the present invention as described in the claims. Moreover, the present invention described herein may have other embodiments and may be implemented or realized in a variety of ways.

Claims

1. A switching power supply auxiliary design method, characterized in that: The method comprises: Build an experimental circuit, select multiple influencing parameters that affect power supply performance, and perform simulation and physical experimental tests based on the influencing parameters and the experimental circuit to obtain data of each influencing parameter under each test; Performing linear regression on each influencing parameter under each test to obtain a regression function of the output voltage ripple and other influencing parameters, and determining whether the regression function meets a preset fitting condition; If the regression function meets the preset fitting condition, the regression coefficients of the output voltage ripple and other influencing parameters are obtained according to the regression function, and causal discovery is performed on any two influencing parameters based on the PC algorithm, and all influencing parameters with causal relationships are obtained according to the causal discovery results, and then the causal effect values ​​corresponding to the influencing parameters and each causal relationship are obtained through the structural equation model; Define each influencing parameter as an entity, summarize the entity relationships between all entities, and associate the corresponding regression coefficients or causal effect values ​​according to the entity relationships to obtain the power supply knowledge graph; The power supply design requirements are obtained, and the parameters to be debugged are searched from the power supply knowledge graph according to the power supply design requirements, and the power supply design scheme is optimized according to the parameters to be debugged.

2. The switching power supply auxiliary design method according to claim 1, characterized in that: The influencing parameters include inductor core material, output capacitance, switching frequency, input voltage, output voltage, output voltage ripple, inductor current ripple, maximum limiting current, and control voltage; The physical experiment test includes a first test experiment and a second test experiment, and the steps of the first test experiment include: Step 1.1, select 220μF output capacitor, select 15μH powder core inductor as inductor core material, set switching frequency to 500kHz, turn on oscilloscope, turn on power supply, set voltage to 12V, and set current limit to 1.5A, turn on electronic load, set constant current mode, and set current to 1.5A; Step 2.1, turn on the power "OUT ON" button, and the waveform of the output voltage Vout displayed on the oscilloscope is a flat horizontal line of 3.3V; Step 3.

1. Turn on the "LOAD ON" button of the electronic load. A triangle waveform with an average value of 1.5A is displayed on the inductor current channel of the oscilloscope, and a flat horizontal line with a level of 3.3V is displayed on the output voltage channel Vout. Step 4.

1. Read the output voltage Vout, measure the output voltage ripple and inductor current ripple, and record these values. Then change the input voltage Vin and load current values ​​and repeat this step. Step 5.

1. Turn off the "LOAD ON" button of the electronic load and the "OUT ON" button of the DC power supply, then connect a 15μH ferrite core inductor and repeat steps 2.1 to 4.

1. Step 6.1, set the switching frequency to 500kHz, and repeat steps 2.1 to 5.1; Step 7.

1. At the end of the measurement, turn off the "LOAD ON" button of the electronic load and the "OUT ON" button of the DC power supply, and then turn off all instruments.

3. The switching power supply auxiliary design method according to claim 1, characterized in that: The steps of the second test experiment include: Step 1.2, select an output capacitor of 220μF, select a powder core inductor with a core material X of 15μH, set the switching frequency to 250kHz, turn on the oscilloscope, turn on the power supply, set the voltage to 12V, and set the current limit to 1.5A, turn on the electronic load, set the constant current mode, and set the current to 1A; Step 2.2, turn on the power "OUT ON" button, and the output voltage waveform displayed on the oscilloscope is a flat horizontal line of 3.3V; Step 3.2, turn on the "LOAD ON" button of the electronic load. At this time, a triangle waveform with an average value of 1A is displayed on the inductor current channel of the oscilloscope, a flat horizontal line with a level between 500mV and 1V and very small ripple is seen on the control voltage channel, and a flat horizontal line with a level of 3.3V is seen on the output voltage channel; Step 4.2, slowly increase the current of the electronic load, and detect the maximum current limit value when the output voltage is stabilized at 3.3V. When the output voltage is no longer stable, turn off the "LOAD ON" button of the electronic load and the "OUT ON" button of the power supply, then record the maximum current limit and the corresponding control voltage value, change the input voltage value, and repeat this step; Step 5.2, select a 15μH ferrite core inductor or a 15μH magnetic powder core inductor, a 10μF output capacitor or a 220μF output capacitor, a 250kHz switching frequency or a 500kHz switching frequency, and repeat steps 2.2 to 4.2 by cross-combining the three; Step 6.

2. At the end of the measurement, turn off the "LOAD ON" button of the electronic load and the "OUT ON" button of the DC power supply, and then turn off all instruments.

4. The switching power supply auxiliary design method according to claim 1, characterized in that: The steps of performing linear regression on each influencing parameter under each test to obtain a regression function of the output voltage ripple and other influencing parameters, and judging whether the regression function meets the preset fitting condition include: Linear regression is performed according to the following formula: ； in, Indicates the inductor core material, Indicates the input voltage, Indicates the maximum limiting current, represents the control voltage, represents the switching frequency, Indicates the output voltage, represents the output voltage ripple, represents the inductor current ripple, represents the output capacitance, represents the intercept term, , , , , , , , are regression coefficients, represents the error term; The preset fitting conditions include whether the multivariate determination coefficient after the regression function is adjusted is greater than a first preset threshold, and whether the p-value of the regression function is greater than a second preset threshold; The adjusted multivariate determination coefficient was calculated according to the following formula: ； in, represents the adjusted multivariate determination coefficient, represents the original multivariate determination coefficient, represents the experimental data capacity, Indicates the number of independent variables; The original multivariate determination coefficient was calculated according to the following formula: ； in, represents the residual sum of squares, represents the total sum of squares, represents the fitting value of the qth experimental data, represents the average value of the experimental data, represents the qth experimental data, and q represents the number of experimental data.

5. The switching power supply auxiliary design method according to claim 1, characterized in that: The steps of performing causal discovery on any two influencing parameters based on the PC algorithm and obtaining all influencing parameters with causal relationships according to the causal discovery results include: Determine each pair of influencing parameters and Whether it is independent under a given set of variables Y: ； If independent, delete and The edges between them are identified, and the directions of the edges are determined by identifying the V-shaped structure and applying the transitivity rule, thus obtaining a directed acyclic graph used to represent the causal relationship between the influencing parameters.

6. The switching power supply auxiliary design method according to claim 5, characterized in that: The step of obtaining the causal effect value corresponding to the influencing parameter and each causal relationship through the structural equation model also includes: The causal effect value is calculated according to the following formula: ； in, , Respectively represent the i-th and j-th influencing parameters, represents the error term, Influencing parameters Influencing parameters The causal effect value of Influencing parameters and influencing parameters The covariance between Influencing parameters The variance of .

7. The switching power supply auxiliary design method according to claim 1, characterized in that: The steps of obtaining power supply design requirements, searching for parameters to be debugged from the power supply knowledge graph according to the power supply design requirements, and optimizing the power supply design solution according to the parameters to be debugged include: Acquire at least one power supply design requirement, each of the power supply design requirements including a target influencing parameter and a value corresponding to the target influencing parameter; The parameters to be debugged that are related to each of the target influencing parameters are searched from the power supply knowledge graph, and the power supply design results are optimized according to the parameters to be debugged to obtain the optimal design solution.

8. A switching power supply auxiliary design system, characterized in that: The system comprises: A simulation and physical experiment platform building module is used to build an experimental circuit, select multiple influencing parameters that affect the performance of the power supply, and perform simulation and physical experiment tests based on the influencing parameters and the experimental circuit to obtain data of each influencing parameter under each test; A linear regression module is used to perform linear regression on each influencing parameter under each test, obtain a regression function of the output voltage ripple and other influencing parameters, and determine whether the regression function meets a preset fitting condition; A causal inference module, for obtaining the regression coefficients of the output voltage ripple and other influencing parameters according to the regression function if the regression function satisfies the preset fitting condition, and performing causal discovery on any two influencing parameters based on the PC algorithm, and obtaining all influencing parameters with causal relationships according to the causal discovery results, and then obtaining the causal effect values ​​corresponding to the influencing parameters and each causal relationship through the structural equation model; The knowledge graph construction module is used to define each influencing parameter as an entity, summarize the entity relationships between all entities, and associate the corresponding regression coefficients or causal effect values ​​according to the entity relationships to obtain the power knowledge graph; The auxiliary design module is used to obtain power supply design requirements, search for parameters to be debugged from the power supply knowledge graph according to the power supply design requirements, and optimize the power supply design solution according to the parameters to be debugged.

9. A storage medium, characterized in that: The storage medium stores one or more programs, which, when executed by a processor, implement the switching power supply auxiliary design method according to any one of claims 1 to 7.

10. An electronic device, characterized in that: The electronic device comprises a memory and a processor, wherein: The memory is used to store computer programs; When the processor is used to execute the computer program stored in the memory, the switching power supply auxiliary design method as described in any one of claims 1 to 7 is implemented.

Images