Transient response optimization method based on linear voltage regulator

Abstract

The present application relates to power management technical field, especially to a kind of transient response optimization method based on linear voltage regulator, method includes: obtaining the transient response data of linear voltage regulator when load mutation, judge whether into transient early warning state.If into early warning state, high-frequency noise threshold segmentation is carried out to output voltage waveform, extract voltage rate of change characteristic, construct compensation efficiency model based on dynamic response time and effective compensation bandwidth, calculate optimal open-loop compensation coefficient, and generate compensation control instruction, drive compensation network to carry out dynamic compensation.If not into early warning state, the ripple uniformity data of output voltage is obtained, determine the deviation matrix of actual ripple distribution surface and ideal ripple distribution plane, judge whether there is local ripple abnormal area.The method improves the transient response speed and stability of linear voltage regulator by dynamic monitoring and intelligent compensation, optimizes high-frequency noise processing and ripple uniformity.

Description

Technical Field

[0001] This invention relates to the field of power management technology, and in particular to a transient response optimization method based on a linear regulator. Background Technology

[0002] Linear regulators play a crucial role in electronic systems, their primary function being to provide a stable and accurate voltage output. In many applications, especially those with extremely high power quality requirements, such as communication equipment, medical instruments, and industrial control systems, the performance of linear regulators directly affects the reliability and stability of the entire system. However, in practical applications, linear regulators often face the challenge of sudden load current changes, which can be caused by various factors, such as system startup, load switching, or sudden task execution. Under conditions of sudden load current changes, traditional linear regulators often struggle to quickly adjust their output voltage, leading to overshoot or undershoot phenomena.

[0003] The above content is only used to help understand the technical solution of the present invention and does not represent an admission that the above content is prior art. Summary of the Invention

[0004] The main objective of this invention is to provide a transient response optimization method based on a linear regulator, which aims to solve the technical problem that traditional linear regulators are prone to output voltage overshoot under load current change scenarios, leading to a decrease in power supply stability.

[0005] To achieve the above objectives, this invention provides a transient response optimization method based on a linear regulator, the method comprising:

[0006] Acquire transient response data of the linear regulator when the load changes abruptly, and determine whether the linear regulator has entered a transient warning state based on the transient response data;

[0007] If a transient warning state is entered, the output voltage waveform of the linear regulator is segmented by a high-frequency noise threshold, and the voltage change rate characteristics of the threshold segmentation region are extracted. A compensation efficiency model based on dynamic response time and effective compensation bandwidth is constructed based on the voltage change rate characteristics, and the optimal open-loop compensation coefficient is calculated based on the compensation efficiency model.

[0008] Based on the optimal open-loop compensation coefficient, a compensation control command is generated to drive the compensation network to dynamically compensate the linear regulator;

[0009] If the transient warning state is not entered, the ripple uniformity data of the output voltage of the linear regulator is obtained, and the deviation matrix between the actual ripple distribution surface and the preset ideal ripple distribution plane is determined based on the ripple uniformity data. Based on the deviation matrix, it is determined whether there are local ripple abnormal areas.

[0010] If there are local ripple anomalies, a regional compensation adjustment signal is generated to control the compensation network to perform directional compensation on the anomaly area until the transient response index of the linear regulator is in the normal operating range. The normal operating range is a state with no transient warning and uniform ripple.

[0011] Optionally, acquiring transient response data of the linear regulator during load changes, and determining whether the linear regulator enters a transient warning state based on the transient response data, includes:

[0012] Acquire transient response data of a linear regulator under load change, determine the output voltage waveform data of the linear regulator based on the transient response data, the voltage waveform data includes a time-domain voltage sequence, and extract edge voltage sampling point information of rising and falling edges in the time-domain voltage sequence;

[0013] The voltage change rate of adjacent sampling points at the rising and falling edges is calculated based on the edge voltage sampling point information, and a slope stability index is constructed based on the voltage change rate of the rising and falling edges.

[0014] The slope stability index is used as a transient response warning criterion. The difference between the transient response warning criterion and the preset transient threshold is used to determine whether the linear regulator has entered a transient warning state.

[0015] Optionally, the slope stability index constructed based on the voltage change rate at the rising and falling edges includes:

[0016] The magnitude and direction of the dynamic response gradient are determined based on the rate of change of voltage at the rising and falling edges.

[0017] The sampling points are grouped according to the gradient direction, and the mean gradient magnitude difference and the mean direction deviation angle of adjacent sampling points in each group are calculated.

[0018] The slope stability index is calculated based on the mean gradient magnitude difference and the mean direction deviation angle, wherein the formula for calculating the stability index is:

[0019]

[0020] In the formula, S is the stability index, and a larger value indicates a more stable slope; ΔG is the mean value of the gradient magnitude difference. This represents the mean directional deviation angle.

[0021] Optionally, the step of constructing a compensation efficiency model based on dynamic response time and effective compensation bandwidth using the voltage change rate characteristics, and calculating the optimal open-loop compensation coefficient based on the compensation efficiency model, includes:

[0022] Based on the voltage change rate characteristics, frequency domain analysis is performed on the compensation region after threshold segmentation to calculate the dynamic response time and effective compensation bandwidth. A compensation factor is then defined based on the dynamic response time and effective compensation bandwidth, with the following formula:

[0023]

[0024] In the formula, As a compensation factor, This refers to the nominal bandwidth of the voltage regulator. For dynamic response time, To effectively compensate for bandwidth;

[0025] A compensation efficiency model is established between the compensation factor and the optimal open-loop compensation coefficient. The compensation efficiency model includes the mapping relationship between the compensation factor and the optimal open-loop compensation coefficient. The optimal open-loop compensation coefficient that satisfies the target response accuracy is calculated based on the mapping relationship.

[0026] The formula for calculating the mapping relationship is as follows:

[0027]

[0028] In the formula, This is the optimal open-loop compensation coefficient. The initial compensation coefficient, The sensitivity coefficient is determined by the voltage regulator's load characteristics and output capacitor parameters. This is a compensation factor.

[0029] Optionally, determining the deviation matrix between the actual ripple distribution surface and the preset ideal ripple distribution plane based on the ripple uniformity data, and determining whether there are local ripple anomaly regions based on the deviation matrix, includes:

[0030] The ripple uniformity data is converted into a frequency domain ripple spectrum matrix, and the actual ripple distribution surface is obtained by fitting.

[0031] The actual ripple distribution surface is registered with the preset ideal ripple plane in the frequency domain, and the ripple amplitude deviation at each frequency point is calculated.

[0032] Based on the ripple amplitude deviation, a deviation matrix is ​​constructed to determine the proportion of frequency points and energy distribution where the absolute value of the deviation exceeds a preset ripple threshold.

[0033] The presence of local ripple anomalies is determined based on the frequency point ratio and energy distribution.

[0034] Optionally, if a local ripple anomaly region exists, generating a region compensation adjustment signal to control the compensation network to perform directional compensation for the anomaly region includes:

[0035] If there are local ripple anomaly regions, a frequency weight matrix is ​​generated based on the deviation matrix, and the weight values ​​are positively correlated with the absolute value of the ripple amplitude deviation.

[0036] Based on the frequency weight matrix, a regional compensation adjustment signal is generated, and the bandpass filter parameters of the compensation network are adjusted. The adjustment intensity is proportional to the gradient value of the weight matrix, and priority is given to directional compensation for high-weight abnormal frequency bands.

[0037] Optionally, the step of generating a compensation control command based on the optimal open-loop compensation coefficient to drive the compensation network to dynamically compensate the linear regulator includes:

[0038] Based on the optimal open-loop compensation coefficient, a compensation control command is generated, and the input voltage fluctuation and load current change are monitored in real time to dynamically adjust the priority of compensation energy allocation.

[0039] Based on the compensation control command and energy allocation priority, the adjustable resistor and capacitor array in the compensation network is driven to achieve dynamic compensation of the loop gain of the linear regulator.

[0040] Optionally, the dynamic adjustment of the compensation energy allocation priority includes:

[0041] When the input voltage fluctuation exceeds the preset voltage amplitude, the high-frequency compensation path is activated first to suppress voltage overshoot. The high-frequency compensation path includes an RC parallel network.

[0042] When the load current change rate exceeds the preset current change rate threshold, the system switches to a low-frequency compensation path to optimize steady-state error and ensure that the compensation energy matches the transient requirements. The low-frequency compensation path includes an LC series network.

[0043] Optionally, performing high-frequency noise threshold segmentation on the output voltage waveform of the linear regulator includes:

[0044] The output voltage waveform of the linear regulator is converted into a frequency domain power spectrum, a high-frequency noise warning threshold is set as a segmentation benchmark, and frequency bands exceeding the benchmark are marked as frequency bands to be compensated.

[0045] The frequency bands to be compensated are grouped according to frequency connectivity. The grouping includes edge noise region and harmonic noise region. The edge noise region adopts RC lead compensation network, and the harmonic noise region adopts LC notch filter for targeted suppression.

[0046] Furthermore, to achieve the above objectives, the present invention also provides a transient response optimization system based on a linear regulator, the transient response optimization system based on a linear regulator comprising:

[0047] The early warning judgment module is used to acquire transient response data of the linear regulator when the load changes suddenly, and to determine whether the linear regulator has entered a transient early warning state based on the transient response data;

[0048] The compensation calculation module is used to perform high-frequency noise threshold segmentation on the output voltage waveform of the linear regulator if a transient warning state is entered, extract the voltage change rate characteristics of the threshold segmentation region, construct a compensation efficiency model based on dynamic response time and effective compensation bandwidth through the voltage change rate characteristics, and calculate the optimal open-loop compensation coefficient according to the compensation efficiency model.

[0049] The dynamic compensation module is used to generate compensation control commands based on the optimal open-loop compensation coefficient, and drive the compensation network to dynamically compensate the linear regulator.

[0050] The ripple analysis module is used to acquire the ripple uniformity data of the output voltage of the linear regulator if the transient warning state is not entered, determine the deviation matrix between the actual ripple distribution surface and the preset ideal ripple distribution plane based on the ripple uniformity data, and determine whether there are local ripple abnormal areas based on the deviation matrix.

[0051] The regional compensation module is used to generate a regional compensation adjustment signal if there is a local ripple abnormality area, and control the compensation network to perform directional compensation for the abnormal area until the transient response index of the linear regulator is in the normal operating range. The normal operating range is a state with no transient warning and uniform ripple.

[0052] This invention provides a transient response optimization method based on a linear voltage regulator. This method acquires transient response data of the linear voltage regulator under load changes and constructs a compensation efficiency model based on voltage change rate characteristics, enabling rapid determination of whether the system has entered a transient warning state. In the transient warning state, by calculating the optimal open-loop compensation coefficient and driving the compensation network for dynamic compensation, overshoot or undershoot of the output voltage is effectively suppressed, significantly improving the speed and stability of the transient response. This dynamic compensation mechanism can quickly adjust the output voltage during load changes, ensuring the continuity and reliability of the system power supply. High-frequency noise threshold segmentation technology is used to extract the voltage change rate characteristics of the threshold segmentation region, enabling accurate identification and separation of high-frequency noise components. By constructing a compensation efficiency model based on dynamic response time and effective compensation bandwidth, the high-frequency noise processing capability is further optimized, reducing the impact of noise on the output voltage and thus improving power supply quality. When not in the transient warning state, the solution analyzes the ripple uniformity data of the output voltage to determine the deviation matrix between the actual ripple distribution surface and the preset ideal ripple distribution plane, accurately identifying local ripple anomaly areas. By generating a regional compensation adjustment signal, directional compensation is performed on abnormal regions, which effectively improves the ripple uniformity of the output voltage and avoids the impact of local ripple anomalies on system performance. Attached Figure Description

[0053] Figure 1 This is a schematic diagram of the transient response optimization device structure based on a linear regulator for the hardware operating environment involved in the embodiments of the present invention;

[0054] Figure 2 This is a flowchart illustrating the first embodiment of the transient response optimization method based on a linear regulator according to the present invention.

[0055] Figure 3 This is a structural block diagram of the first embodiment of the transient response optimization system based on a linear voltage regulator according to the present invention.

[0056] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0057] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0058] Reference Figure 1 , Figure 1 This is a schematic diagram of the transient response optimization device based on a linear regulator for the hardware operating environment involved in the embodiments of the present invention.

[0059] like Figure 1As shown, the transient response optimization device based on a linear regulator may include: a processor 1001, such as a central processing unit (CPU), a communication bus 1002, a user interface 1003, a network interface 1004, and a memory 1005. The communication bus 1002 is used to enable communication between these components. The user interface 1003 may include a display screen, and optionally, it may also include a standard wired interface or a wireless interface. In this invention, the wired interface of the user interface 1003 may be a USB interface. The network interface 1004 may optionally include a standard wired interface or a wireless interface (such as a Wi-Fi interface). The memory 1005 may be a high-speed random access memory (RAM) or a non-volatile memory (NVM), such as a disk storage device. The memory 1005 may also optionally be a storage device independent of the aforementioned processor 1001.

[0060] Those skilled in the art will understand that Figure 1 The structure shown does not constitute a limitation on the transient response optimization device based on the linear regulator, and may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0061] like Figure 1 As shown, the memory 1005, which serves as a computer storage medium, may include an operating system, a network communication module, a user interface module, and a transient response optimization program based on a linear regulator.

[0062] exist Figure 1 In the transient response optimization device based on a linear regulator shown, the network interface 1004 is mainly used to connect to the backend server and communicate with the backend server; the user interface 1003 is mainly used to connect to peripherals; the transient response optimization device based on a linear regulator calls the transient response optimization program based on a linear regulator stored in the memory 1005 through the processor 1001 and executes the transient response optimization method based on a linear regulator provided in this embodiment of the invention.

[0063] Based on the above hardware structure, an embodiment of the transient response optimization method based on a linear regulator of the present invention is proposed.

[0064] Reference Figure 2 , Figure 2 This is a flowchart illustrating the first embodiment of the transient response optimization method based on a linear voltage regulator according to the present invention. The first embodiment of the transient response optimization method based on a linear voltage regulator according to the present invention is presented.

[0065] In the first embodiment, the transient response optimization method based on a linear regulator includes the following steps:

[0066] Step S100: Obtain transient response data of the linear regulator when the load changes suddenly, and determine whether the linear regulator has entered the transient warning state based on the transient response data.

[0067] Transient response data can be waveform data of the output voltage of a linear regulator changing over time when the load current changes abruptly, including parameters such as voltage overshoot, undershoot, and recovery time. This data can be obtained by acquiring the output voltage signal in real time and recording the time series. A transient warning state can be an abnormal state where the system output voltage deviates from the target value by more than a preset percentage or the recovery time exceeds a threshold. In one specific embodiment, the determination process can be achieved by comparing the difference between the actual voltage and the target value, or by calculating whether the recovery time parameter exceeds a preset standard. For example, a warning is triggered when the voltage deviation exceeds ±5% or the recovery time exceeds 1μs. This technology provides a basis for subsequent compensation decisions by monitoring voltage fluctuation characteristics in real time.

[0068] Step S200: If the transient warning state is entered, the high-frequency noise threshold segmentation is performed on the output voltage waveform of the linear regulator, the voltage change rate characteristics of the threshold segmentation region are extracted, a compensation efficiency model based on dynamic response time and effective compensation bandwidth is constructed based on the voltage change rate characteristics, and the optimal open-loop compensation coefficient is calculated based on the compensation efficiency model.

[0069] High-frequency noise threshold segmentation can be a technique that divides a waveform into high-frequency noise regions and stable regions by setting a voltage change rate or amplitude threshold. For example, the interval where the voltage change rate exceeds 2V / μs can be set as a high-frequency noise region, or high-frequency components can be separated by wavelet transform. The voltage change rate feature can be the derivative of the voltage over time within the segmented region, such as the instantaneous slope or rate of change, which can be obtained by calculating the voltage difference between adjacent sampling points and dividing by the time interval. The compensation efficiency model can be a mathematical expression combining dynamic response time and effective compensation bandwidth, where dynamic response time is the time required for the system to go from abrupt change to stability, and effective compensation bandwidth is the frequency range that the compensation network can handle, for example, by establishing parameter correlation through transfer functions or optimization algorithms. In a specific embodiment, the model may aim to minimize overshoot / undershoot amplitude or shorten recovery time, and obtain the optimal open-loop compensation coefficient through iterative calculation. This technique ensures that the compensation effect matches the dynamic characteristics of the system by quantifying the degree of noise interference and optimizing the compensation parameters.

[0070] Step S300: Generate compensation control commands based on the optimal open-loop compensation coefficients to drive the compensation network to dynamically compensate the linear regulator.

[0071] Compensation control commands can be parameter configurations that convert compensation coefficients into analog or digital control signals, such as adjusting the error amplifier gain, adjustable capacitor values, or feedback loop parameters. The compensation network can include adjustable components such as capacitor arrays, resistors, or operational amplifiers, whose parameters can be adjusted in real time via digital signals or analog circuitry. Dynamic compensation can be achieved by changing the impedance characteristics of the compensation network, for example, increasing the compensation capacitor to improve low-frequency response or adjusting the feedback loop gain to optimize high-frequency response. In one specific embodiment, the control command may drive a digital potentiometer via a pulse-width modulation (PWM) signal to change the equivalent resistance value of the compensation network in real time. This technique operates by rapidly adjusting system parameters to offset voltage fluctuations caused by sudden load changes, suppressing overshoot or undershoot.

[0072] Step S400: If the transient warning state is not entered, obtain the ripple uniformity data of the output voltage of the linear regulator, determine the deviation matrix between the actual ripple distribution surface and the preset ideal ripple distribution plane based on the ripple uniformity data, and determine whether there is a local ripple abnormal area based on the deviation matrix.

[0073] Ripple uniformity data can be the distribution characteristics of high-frequency fluctuations in the output voltage of a voltage regulator, which can be obtained through Fourier transform analysis of frequency domain components or time-domain statistical methods. The actual ripple distribution surface can be a three-dimensional data structure describing the ripple amplitude at different frequencies or time points, while the preset ideal ripple distribution plane can be the desired uniform amplitude distribution benchmark. The deviation matrix can be an error matrix obtained by comparing the actual and ideal distributions, for example, by using matrix subtraction or statistical error indices to quantify the amplitude differences in each region. In a specific embodiment, the deviation matrix can be generated as a two-dimensional error distribution map by calculating the sum of squares of the differences between the actual and ideal ripple amplitudes. This technique provides a basis for subsequent compensation by accurately identifying the non-uniformity of the ripple distribution.

[0074] In step S500, if there is a local ripple abnormality area, a regional compensation adjustment signal is generated to control the compensation network to perform directional compensation for the abnormal area until the transient response index of the linear regulator is in the normal operating range. The normal operating range is a state with no transient warning and uniform ripple.

[0075] The regional compensation adjustment signal can be a control command targeting a specific frequency band or time window, such as suppressing ripple at a specific frequency through an adjustable filter or adjusting the local compensation intensity through pulse width modulation. The compensation network can have spatial or frequency domain selectivity, for example, employing a multi-channel adjustable capacitor array or a digitally programmable filter to adjust the compensation parameters for the corresponding region based on the signal. In one specific embodiment, regional compensation may reduce the ripple amplitude in abnormal regions by dynamically adjusting the compensation capacitor value for a specific frequency band. This technique improves ripple uniformity by precisely locating and correcting abnormal regions, avoiding local overshoot or noise accumulation that could affect system stability.

[0076] This embodiment provides a transient response optimization method based on a linear regulator. By monitoring transient response data during load surges in real time and determining the warning state, it calculates the optimal compensation coefficient using high-frequency noise threshold segmentation and a compensation efficiency model, driving the compensation network to dynamically adjust system parameters. In the non-warning state, it generates a directional compensation signal by analyzing the ripple distribution deviation matrix, controlling the compensation network to correct local abnormal regions. This method employs a scenario-specific dynamic compensation strategy to quickly suppress voltage overshoot / undershoot and shorten recovery time during transient warnings, while eliminating ripple distribution non-uniformity in steady-state conditions. Ultimately, it achieves high response speed and output stability of the linear regulator under both load surges and steady-state operating modes, while avoiding resource waste caused by overcompensation, significantly improving the reliability and energy efficiency of the power management system.

[0077] In one embodiment, the transient response data of the linear regulator during a sudden load change is acquired, and a determination is made based on the transient response data whether the linear regulator has entered a transient warning state, including:

[0078] Acquire transient response data of the linear regulator under load change, determine the output voltage waveform data of the linear regulator based on the transient response data, the voltage waveform data includes the time-domain voltage sequence, and extract the edge voltage sampling point information of the rising and falling edges in the time-domain voltage sequence;

[0079] The transient response data is a time-domain record of the output voltage of the linear regulator under sudden load changes, which can be obtained through real-time sampling or storage. The output voltage waveform data is a continuous sampling sequence reflecting the voltage change over time within the transient response data; its time-domain voltage sequence can be generated by an analog-to-digital converter at a preset sampling frequency. Edge voltage sampling point information specifically refers to discrete sampling point data in the rising and falling edge regions of the voltage waveform, including the voltage value and timestamp at the corresponding moment, which can be identified using threshold comparison or edge detection algorithms. For example, when the voltage difference between adjacent sampling points exceeds a preset threshold, it can be marked as the starting point of an edge region, and continuous sampling point data within that region can be extracted.

[0080] The voltage change rate of adjacent sampling points at the rising and falling edges is calculated based on edge voltage sampling point information, and a slope stability index is constructed based on the voltage change rate of the rising and falling edges.

[0081] The rate of change of voltage is the ratio of the voltage value between adjacent sampling points to the time interval, and its magnitude characterizes the steepness of the voltage waveform in the time domain. For example, the rate of change of voltage at adjacent sampling points can be calculated point-by-point using the finite difference method, or a sliding window averaging method can be used to reduce noise interference. The slope stability index is a quantitative parameter that combines the statistical characteristics of the rate of change of voltage at both the rising and falling edges.

[0082] The slope stability index is used as the transient response warning criterion. The difference between the transient response warning criterion and the preset transient threshold is used to determine whether the linear regulator enters the transient warning state.

[0083] The preset transient threshold is a critical value determined based on historical data or theoretical models, used to distinguish between normal and abnormal transient responses. For example, it can be determined by comparing the absolute value of the difference between a stability index and the threshold to see if it exceeds a preset range, or by judging whether the index has consistently exceeded the threshold for a certain period. Dynamic threshold adaptability can be achieved by adjusting the threshold parameters; for example, lowering the threshold in high-frequency load switching scenarios to provide early warning, and raising the threshold in low-frequency scenarios to reduce false alarms. This comparison process, by quantifying the voltage change characteristics of edge regions, can accurately identify voltage abrupt anomalies. Compared to traditional voltage deviation threshold comparison methods, it can capture potential fluctuation trends earlier and reduce the probability of false triggering caused by noise interference.

[0084] This embodiment acquires transient response data during load surges and extracts voltage waveform edge features. It calculates the voltage change rate to construct a slope stability index, which is then compared with a dynamic threshold to determine the warning status. This achieves the following technical effects: focusing on the critical region of voltage surges to improve feature extraction accuracy; combining multi-dimensional statistical parameters to construct a comprehensive stability index to fully reflect transient response quality; and adjusting the dynamic threshold to adapt to different load scenarios. These techniques work synergistically to significantly enhance the accuracy and timeliness of transient warnings, shorten warning delay time, and provide a reliable basis for subsequent compensation strategies, thereby effectively improving the stability and response speed of linear regulators under load surge scenarios.

[0085] In one embodiment, a slope stability index is constructed based on the voltage change rate at the rising and falling edges, including: determining the dynamic response gradient magnitude and gradient direction based on the voltage change rate at the rising and falling edges; grouping the sampling points according to the gradient direction, calculating the mean difference in gradient magnitude and the mean angle of directional deviation between adjacent sampling points in each group; and calculating the slope stability index based on the mean difference in gradient magnitude and the mean angle of directional deviation, wherein the calculation formula for the stability index is:

[0086]

[0087] In the formula, S is the stability index, and a larger value indicates a more stable slope; ΔG is the mean value of the gradient magnitude difference. This represents the mean directional deviation angle.

[0088] The magnitude of the dynamic response gradient can be the absolute value of the rate of voltage change, i.e., the absolute value of the ratio of the voltage change between adjacent sampling points to the time interval. For example, this value can reflect the drastic nature of the voltage change. The gradient direction can be a directional attribute of the voltage change, which can be quantified by a sign (e.g., a positive sign represents the rising edge, and a negative sign represents the falling edge) or a vector angle. For example, if the voltage at a sampling point rises from 1V to 1.2V, then its gradient direction is upward. By calculating the gradient magnitude and direction of the edge voltage sampling point sequence point by point, the voltage change can be decomposed into two independent dimensions: magnitude and direction, providing structured data support for subsequent analysis.

[0089] Grouping sampling points based on gradient direction can be an operation that classifies them according to direction sign or angle range. For example, rising edge and falling edge sampling points can be assigned to different groups, or the groups can be further subdivided according to angle intervals. By comparing the gradient direction sign or angle difference between adjacent sampling points, if the direction deviation angle is less than a preset threshold, they are considered to be in the same direction group. This operation can eliminate the interference of direction changes on amplitude analysis, ensure the consistency of data in the same group in terms of directional characteristics, and thus improve the accuracy of subsequent calculations. The average gradient amplitude difference can be the average of the absolute values ​​of the gradient amplitude differences between adjacent sampling points in the same group. The average direction deviation angle can be the average of the absolute values ​​of the gradient direction angle differences between adjacent sampling points in the same group. By calculating the amplitude difference and direction angle difference of adjacent sampling points pair by pair and taking the average, the amplitude fluctuation and direction deviation of voltage changes in the same direction group can be quantified, thereby reflecting the local stability of the slope.

[0090] In one embodiment, a compensation efficiency model based on dynamic response time and effective compensation bandwidth is constructed using voltage change rate characteristics. The optimal open-loop compensation coefficient is then calculated based on the compensation efficiency model, including:

[0091] Based on the voltage change rate characteristics, frequency domain analysis is performed on the compensation region after threshold segmentation to calculate the dynamic response time and effective compensation bandwidth. A compensation factor is then defined based on the dynamic response time and effective compensation bandwidth, with the following formula:

[0092]

[0093] In the formula, As a compensation factor, This refers to the nominal bandwidth of the voltage regulator. For dynamic response time, To effectively compensate for bandwidth;

[0094] A compensation efficiency model is established between the compensation factor and the optimal open-loop compensation coefficient. The compensation efficiency model includes the mapping relationship between the compensation factor and the optimal open-loop compensation coefficient. The optimal open-loop compensation coefficient that satisfies the target response accuracy is calculated based on the mapping relationship.

[0095] The formula for calculating the mapping relationship is as follows:

[0096]

[0097] In the formula, This is the optimal open-loop compensation coefficient. The initial compensation coefficient, The sensitivity coefficient is determined by the voltage regulator's load characteristics and output capacitor parameters. This is a compensation factor.

[0098] The compensation factor is a parameter that quantifies the transient response efficiency of a compensation network. Its definition is based on the effective compensation bandwidth, dynamic response time, and the nominal bandwidth of the voltage regulator. The effective compensation bandwidth is the upper limit of the high-frequency noise suppression frequency that the compensation network can handle, and it can be determined by configuring the parameters of the compensation network components. The dynamic response time is the time required for the linear voltage regulator to recover to the percentage of the target value from a load change, for example, calculated by monitoring the time window when the voltage waveform recovers to 90% of its stable value. The nominal bandwidth of the voltage regulator is its inherent frequency response capability under no load change, determined by the internal amplifier gain and feedback loop design. Frequency domain analysis techniques, such as Fast Fourier Transform (FFT) or wavelet analysis on the high-frequency noise region after threshold segmentation, can be used to extract the spectral characteristics of the voltage waveform, and then calculate the above parameters. The calculation formula for the compensation factor combines the system's dynamic characteristics and inherent parameters; a larger value indicates a stronger response capability of the compensation network under the current load change condition, thus providing a quantitative basis for subsequent adjustment of the compensation coefficient.

[0099] The optimal open-loop compensation coefficient is a parameter determined by a linear combination of the compensation factor and the initial compensation coefficient. The initial compensation coefficient is the baseline parameter of the compensation network under steady-state conditions, usually preset according to the voltage regulator design objectives to ensure stable output under normal operating conditions. The compensation sensitivity coefficient is a weighted coefficient reflecting the influence of load characteristics and output capacitance parameters on compensation requirements, and can be obtained by real-time monitoring of load current values ​​and querying a pre-stored parameter table or by online calculation.

[0100] By substituting the compensation factor, initial compensation coefficient, and compensation sensitivity coefficient into the mapping formula, the optimal open-loop compensation coefficient can be calculated in real time. This mapping relationship, through linear combination, retains the steady-state reference compensation capability while dynamically adjusting the compensation intensity according to transient requirements. For example, it increases the compensation intensity under high impedance loads or high ESR capacitance conditions, thus avoiding the limitations of fixed compensation strategies. The technical operation includes frequency domain analysis of the compensation region after threshold segmentation, compensation factor quantization, determination of compensation sensitivity coefficient, and solution of the optimal compensation coefficient. Frequency domain analysis can be achieved through FFT or wavelet analysis, for example, by identifying the voltage recovery time window to determine the dynamic response time, or by determining the effective compensation bandwidth through the spectral cutoff frequency. Compensation factor quantization is completed by substituting parameters into the formula, while the determination of the compensation sensitivity coefficient can be achieved by querying a pre-stored table or an online calculation function based on real-time load parameters. Finally, all parameters are integrated through the linear mapping formula to output the optimal open-loop compensation coefficient in real time. This process avoids the delay of complex iterative algorithms, ensuring that the compensation network can respond to load changes within microseconds.

[0101] This implementation correlates dynamic response time, effective compensation bandwidth, and the nominal bandwidth of the voltage regulator through a compensation factor, achieving precise quantification of compensation requirements. Dynamic adjustment of the compensation sensitivity coefficient enhances the system's adaptability to different loads and capacitance parameters. Linear mapping simplifies the calculation process, ensuring that the compensation coefficient is generated rapidly within microseconds. These improvements work synergistically to significantly enhance the accuracy and real-time performance of transient response, reduce voltage overshoot / undershoot, shorten recovery time, and expand the method's versatility for different voltage regulator models or application scenarios, ultimately achieving higher quality voltage output stability.

[0102] In one embodiment, the deviation matrix between the actual ripple distribution surface and the preset ideal ripple distribution plane is determined based on ripple uniformity data. The presence of local ripple anomaly regions is then determined based on the deviation matrix, including:

[0103] The ripple uniformity data is converted into a frequency domain ripple spectrum matrix, and the actual ripple distribution surface is obtained by fitting.

[0104] The actual ripple distribution surface is registered with the preset ideal ripple plane in the frequency domain, and the ripple amplitude deviation at each frequency point is calculated.

[0105] Based on the ripple amplitude deviation, a deviation matrix is ​​constructed to determine the proportion of frequency points and energy distribution where the absolute value of the deviation exceeds the preset ripple threshold.

[0106] The presence of local ripple anomalies can be determined based on the frequency point ratio and energy distribution.

[0107] The ripple uniformity data can be a time-domain signal characterizing the output voltage fluctuation of a linear regulator, converted into a frequency-domain ripple spectrum matrix using Fourier transform or other frequency-domain analysis methods, including but not limited to Fast Fourier Transform, Wavelet Transform, and Short-Time Fourier Transform. The rows or columns of this matrix can correspond to different frequency points, with each element representing the ripple amplitude at that frequency. Its purpose is to map time-domain fluctuation characteristics to the frequency domain, facilitating the quantification of the energy distribution of different frequency components. The actual ripple distribution surface can be a three-dimensional surface model, with the horizontal axis representing frequency, the vertical axis representing time or spatial location, and the height representing the ripple amplitude. It is generated based on the frequency-domain ripple spectrum matrix using a surface fitting algorithm, such as least squares or Gaussian process regression. The preset ideal ripple plane can be a flat amplitude distribution benchmark desired in the design objectives, typically a two-dimensional plane with uniform amplitude below a preset threshold.

[0108] Frequency domain registration can be achieved through interpolation or phase correction algorithms. For example, by adjusting the frequency resolution or eliminating phase shifts caused by signal delay, the frequency coordinates of the actual ripple distribution surface are aligned with the frequency coordinates of a preset ideal plane. Ripple amplitude deviation can be the difference in amplitude between the actual and ideal distribution at corresponding frequency points, including absolute or relative deviations, calculated by point-by-point subtraction. The deviation matrix can be a structured data matrix, storing deviation values ​​with frequency points as indices. Its construction must retain frequency coordinate information, such as using row or column indices to correspond to specific frequency values. The frequency point percentage can be the proportion of frequency points with absolute deviations exceeding a preset threshold to the total number of frequency points. The energy distribution can be the proportion of the sum of squared ripple amplitudes at abnormal frequency points to the total energy, statistically calculated by determining the number of frequency points meeting the conditions and the total energy.

[0109] In one specific embodiment, the ripple uniformity data is first segmented. Each segment is then subjected to a Fast Fourier Transform (FFT) to generate frequency-amplitude pairs, which are subsequently integrated into a frequency-domain ripple spectrum matrix. This matrix is ​​fitted using the least squares method to generate the actual ripple distribution surface. Simultaneously, a pre-defined ideal plane is interpolated to align the frequency resolution with the actual surface. The amplitude difference between the registered surface and the plane is calculated point-by-point to construct a deviation matrix and mark abnormal frequency points. After statistically analyzing the proportion of abnormal frequency points and energy distribution, if any indicator exceeds a threshold, such as a proportion > 5% or an energy > 10%, a local ripple anomaly region is determined to exist.

[0110] This embodiment converts ripple uniformity data into a frequency domain ripple spectrum matrix and fits it to a surface. Frequency domain registration technology is used to eliminate coordinate deviations, and multi-dimensional statistical indicators are combined to comprehensively determine abnormal regions. These indicators include frequency point proportions and energy distribution, achieving the technical effect of improving the accuracy and reliability of local ripple anomaly detection. This scheme enhances the ability to locate specific frequency components through frequency domain analysis, reducing the risk of misjudgment due to coordinate misalignment. Furthermore, the collaborative evaluation of multiple indicators avoids the limitations of single parameters, ultimately providing accurate frequency coordinates and deviation values ​​for directional compensation, thereby optimizing the output quality of the linear regulator.

[0111] In one embodiment, if a local ripple anomaly region exists, a region compensation adjustment signal is generated to control the compensation network to perform directional compensation for the anomaly region. This includes: if a local ripple anomaly region exists, a frequency weight matrix is ​​generated based on the deviation matrix, where the weight value is positively correlated with the absolute value of the ripple amplitude deviation; a region compensation adjustment signal is generated based on the frequency weight matrix to adjust the bandpass filter parameters of the compensation network, where the adjustment intensity is proportional to the gradient value of the weight matrix, and directional compensation is preferentially performed on high-weight anomaly frequency bands.

[0112] The frequency weight matrix quantifies the compensation priority of different frequency bands. It is obtained by mapping the ripple amplitude deviation data in the deviation matrix to the frequency domain. This matrix can be implemented using frequency domain decomposition techniques such as Fourier transform or wavelet analysis. After converting the time-domain ripple data into a frequency amplitude distribution map, weight values ​​are generated using normalization or weighting functions. The weight values ​​are positively correlated with the absolute value of the ripple amplitude deviation. For example, using a linear proportional or exponential function, when the ripple amplitude of a certain frequency band deviates from the ideal value by 20%, its weight can be set to 2, and when it deviates by 10%, the weight is 1. The construction of this matrix needs to be combined with the frequency resolution requirements, dividing the ripple data into multiple frequency bands and calculating the average deviation of each band.

[0113] Bandpass filter parameters are adjustable parameters in a compensation network used to selectively adjust signals in specific frequency bands. These parameters include the center frequency, bandwidth, and gain. For example, if the ripple amplitude in a certain frequency band is too high, the gain of the bandpass filter in that band can be adjusted to a negative value to cancel the noise; if the amplitude is too low, the gain can be increased to balance the distribution. Parameter adjustment must be matched to the hardware structure of the compensation network, such as using a variable capacitor array or a digitally programmable filter to achieve dynamic adjustment. Adjustments to these parameters need to be fed back to the adjustable components of the compensation network in real time, such as switching the capacitor array or adjusting the feedback resistor of the operational amplifier through a digital control circuit.

[0114] The gradient value of the weight matrix reflects the rate of change of weights between adjacent frequency points, and is calculated using the finite difference method or differential operations. For example, if the weight of a certain frequency band jumps from 1 to 5, the gradient value of that region is large, indicating a drastic change in ripple anomaly. The sign and magnitude of the gradient value determine the distribution direction and intensity of the compensation strength; for example, regions with abrupt weight changes may require stronger compensation intervention. The process of generating the frequency weight matrix involves mapping the ripple amplitude deviation data in the deviation matrix to the frequency domain and determining the weight values ​​using linear proportions or exponential functions. This process requires dividing the frequency bands according to frequency resolution requirements and calculating the average deviation of each band to form a complete frequency weight matrix.

[0115] The process of calculating the gradient value of the weight matrix involves performing gradient operations along the frequency axis to identify regions with drastic weight changes. For example, when the weights of two adjacent frequency points are 3 and 1 respectively, the gradient value is 2, indicating that the ripple anomaly in that region has decreased significantly. The calculation of the gradient value provides a basis for subsequent compensation intensity allocation. The process of generating the regional compensation adjustment signal combines the weight matrix and gradient value, and is implemented through instructions containing frequency band location information and compensation intensity parameters. For example, for frequency bands with high weights and steep gradients, the generated instructions require the bandpass filter to increase the compensation intensity in that band, such as increasing the attenuation or adjusting the phase; for low-weight regions, intervention is reduced. This signal must ensure that compensation resources are preferentially allocated to the frequency bands with the most severe problems. The process of dynamically adjusting the bandpass filter parameters translates the adjustment signal into specific parameter adjustment instructions. For example, if a frequency band has a weight of 5 and a gradient value of 3, it may be possible to lock the center frequency to the abnormal frequency, expand the bandwidth to cover adjacent affected areas, and set the gain to 5dB to offset the ripple amplitude. Parameter adjustments need to be fed back to the adjustable components of the compensation network in real time, such as switching the capacitor array or adjusting the feedback resistor of the operational amplifier through digital control circuitry.

[0116] This embodiment maps ripple amplitude deviation data into a frequency weight matrix to quantify compensation priority, combines gradient value analysis to determine the compensation intensity distribution, and then achieves directional compensation by dynamically adjusting the bandpass filter parameters. This significantly improves the targeting of compensation and the utilization rate of resources. Through frequency domain weight analysis and gradient-driven parameter adjustment, this scheme prioritizes addressing issues in high-weight abnormal frequency bands while avoiding interference with normal signal components. Ultimately, it achieves high accuracy and stability of the output voltage under both load surges and steady-state operation.

[0117] In one embodiment, a compensation control command is generated based on the optimal open-loop compensation coefficient to drive the compensation network to dynamically compensate the linear regulator, including:

[0118] The compensation control command is generated based on the optimal open-loop compensation coefficient, and the input voltage fluctuation and load current change are monitored in real time to dynamically adjust the priority of compensation energy distribution.

[0119] Dynamic compensation of the loop gain of the linear regulator is achieved by using the adjustable resistor-capacitor array in the compensation network driven by compensation control commands and energy allocation priority.

[0120] Input voltage fluctuations can be unexpected fluctuations in the power input voltage over time, which can be obtained by real-time sampling of the input voltage value through a voltage detection circuit. Examples include voltage changes caused by grid interference or upstream power instability. Load current changes can be instantaneous increases or decreases in load current, which can be obtained by acquiring load current values ​​through current sensing resistors or Hall effect sensors. Examples include current surges caused by system module startup or shutdown. Compensation energy allocation priority can be the order or weight allocation of energy calls by different compensation paths or components in the compensation network. Examples include dynamic weight matrices calculated based on preset weights or machine learning models. Real-time monitoring can ensure the timing consistency of voltage and current data through synchronous sampling technology. For example, a high-frequency sampling rate analog-to-digital converter (ADC) is used to convert analog signals into digital signals before transmitting them to the control unit. Dynamic adjustment can be achieved through a priority evaluation algorithm. For example, the input voltage fluctuation amplitude and load current change rate are input into the algorithm to generate a dynamic weight matrix, thereby optimizing the allocation strategy of compensation resources. For instance, when the input voltage drops sharply and the load current surges, the system may prioritize activating the capacitor array to quickly replenish energy while reducing dependence on the resistor network to avoid overcompensation.

[0121] An adjustable resistor-capacitor array can be a structure composed of multiple programmable resistors and capacitors, exemplified by components such as digital potentiometers, switched resistor networks, switched capacitor arrays, or varactor diodes. The loop gain can be the product of the open-loop gain and the feedback coefficient in a feedback control system, exemplified by changing its value by adjusting the impedance characteristics of the compensation network. Driving can be achieved by adjusting the array parameter combination via digital control signals. For example, in high-frequency noise suppression scenarios, increasing the equivalent capacitance of the capacitor array can improve low-frequency compensation capability; while under fast response requirements, decreasing the resistance value can accelerate the loop response speed. This process, by changing the transfer function of the compensation network, multiplies the compensation network by the open-loop transfer function of the regulator's main loop to achieve the target closed-loop gain characteristic, thereby optimizing the balance between system stability and response speed.

[0122] This embodiment generates compensation control commands by real-time monitoring of input voltage fluctuations and load current changes. Combined with an algorithm strategy that dynamically adjusts the priority of compensation energy allocation, it drives an adjustable resistor-capacitor array to change the impedance characteristics of the compensation network, ultimately achieving dynamic compensation of the loop gain. Real-time monitoring ensures the compensation strategy matches the current operating conditions; dynamic priority adjustment avoids overload or energy waste of single components; the adjustable resistor-capacitor array provides discrete or continuous parameter adjustment capabilities; and dynamic adjustment of the loop gain directly optimizes the transient response characteristics of the regulator. These techniques work synergistically to significantly improve the power supply quality of the linear regulator in complex dynamic environments, especially in scenarios with high-frequency noise and sudden load changes. It can more accurately suppress voltage fluctuations, maintain the stability and accuracy of the output voltage, while reducing redundant component configurations and lowering hardware implementation costs.

[0123] In one embodiment, dynamically adjusting the priority of compensation energy allocation includes:

[0124] When the input voltage fluctuation exceeds the preset voltage amplitude, the high-frequency compensation path is activated first to suppress voltage overshoot. The high-frequency compensation path includes an RC parallel network.

[0125] The input voltage fluctuation amplitude can be an electrical parameter characterizing the degree to which the input voltage deviates from the rated value. It can be obtained by real-time sampling and calculation through a voltage monitoring circuit. For example, the input voltage fluctuation amplitude can include the difference between the peak voltage and the rated value or the deviation of the effective value. The preset voltage amplitude can be a threshold parameter used to determine high-frequency disturbance conditions. It can be preset through historical data statistics or simulation analysis. For example, the preset voltage amplitude can include ±5% of the rated voltage range. The high-frequency compensation path can be a compensation circuit structure containing an RC parallel network, where the network composed of resistors and capacitors in parallel has high-frequency and low-impedance characteristics. For example, the high-frequency compensation path can include a parallel circuit composed of a film capacitor and a power resistor. Enabling the high-frequency compensation path can be a control action of conducting the RC parallel path through a switching network. For example, the path switching can be achieved by driving a MOSFET switch by outputting a PWM signal from a digital controller.

[0126] When the load current change rate exceeds the preset current change rate threshold, the system switches to the low-frequency compensation path to optimize steady-state error and ensure that the compensation energy matches the transient demand. The low-frequency compensation path includes an LC series network.

[0127] The load current change rate can be an electrical parameter characterizing the rate of change of load current over time. It can be calculated in real time using a current sensor and a differentiating circuit. For example, the load current change rate can include the increment or derivative of the current amplitude per unit time. The preset current change rate threshold can be a criterion for determining low-frequency load abrupt changes in operating conditions. It can be dynamically adjusted through experimental testing or an adaptive algorithm. The low-frequency compensation path can be a compensation circuit structure containing an LC series network, where the network formed by the series connection of inductors and capacitors has low-frequency resonance characteristics. For example, the low-frequency compensation path can include a series circuit composed of ferrite inductors and electrolytic capacitors. Switching to the low-frequency compensation path can be a control action that disconnects the high-frequency path and connects the LC series path through a switching network. For example, the path switching can be achieved by controlling multiple sets of relays to switch synchronously through state machine logic.

[0128] In terms of technical operation, the path switching mechanism acquires input voltage fluctuation amplitude and load current change rate data through a real-time monitoring module and compares them with preset thresholds. When the input voltage fluctuation exceeds the threshold, the system suppresses overshoot through the rapid charging and discharging characteristics of the high-frequency compensation path; for example, the capacitor in the RC parallel network absorbs instantaneous energy by reducing high-frequency impedance. When the load current change rate exceeds the threshold, the system compensates for steady-state error through the energy storage characteristics of the low-frequency compensation path; for example, the inductor in the LC series network delays voltage drop by storing energy. During the path switching process, the state of the switching elements is synchronously adjusted through digital control signals. For example, when switching to the low-frequency path, the switch of the RC path is opened first and then the switch of the LC path is closed to avoid energy interruption or oscillation during the switching process.

[0129] This embodiment monitors the input voltage fluctuation amplitude and load current change rate in real time, and dynamically activates a high-frequency RC parallel network or a low-frequency LC series network based on preset thresholds. It utilizes the rapid charging and discharging characteristics of the high-frequency path to suppress voltage overshoot, and the energy storage and resonance characteristics of the low-frequency path to compensate for steady-state errors. Simultaneously, synchronous control of the switching network achieves interference-free path switching. This achieves targeted matching of the compensation path according to the disturbance frequency characteristics, significantly improving the efficiency of high-frequency overshoot suppression and the accuracy of low-frequency error compensation. Through a frequency-band compensation strategy and intelligent switching mechanism, this scheme reduces redundant configuration of compensation components, minimizes energy waste, and achieves rapid and stable recovery of the output voltage under complex dynamic conditions.

[0130] In one embodiment, high-frequency noise thresholding is performed on the output voltage waveform of the linear regulator, including:

[0131] The output voltage waveform of the linear regulator is converted into a frequency domain power spectrum. A high-frequency noise warning threshold is set as a segmentation reference, and frequency bands exceeding the reference are marked as frequency bands to be compensated.

[0132] The frequency domain power spectrum can be a mathematical expression of a time-domain signal decomposed into a superposition of sinusoidal waves using Fourier transform or Fast Fourier transform (FFT). It characterizes the frequency distribution features of the signal by quantifying the energy intensity of different frequency components. This can be achieved through digital signal processing algorithms or dedicated spectrum analysis tools. For example, the frequency domain power spectrum can include spectrum analysis results based on Discrete Fourier Transform (DFT) or time-frequency analysis results based on Short-Time Fourier Transform (STFT). The high-frequency noise warning threshold can be a benchmark parameter used to distinguish between significant and acceptable noise interference frequency bands. It can be determined through statistical analysis methods or system preset standards. For example, the threshold can include a dynamic threshold of the noise floor mean plus three standard deviations or a fixed preset value such as 30 dBm. The frequency bands to be compensated can be a set of frequency intervals whose power spectrum amplitude exceeds a set threshold. These frequency bands are selected from the frequency domain power spectrum through threshold segmentation to identify noise frequency bands that require targeted suppression. For example, this can be achieved by: first, inputting the voltage waveform into the FFT module for frequency domain conversion; then, calculating the power spectrum amplitude at each frequency point; next, comparing the amplitude with a preset threshold point by point; and finally, marking all continuous or discrete frequency intervals exceeding the threshold as the frequency bands to be compensated. This achieves the technical effect of accurately locating the distribution range of high-frequency noise and narrowing the subsequent compensation design range.

[0133] The frequency bands to be compensated are grouped according to frequency connectivity. The groups include edge noise region and harmonic noise region. The edge noise region adopts RC lead compensation network, and the harmonic noise region adopts LC notch filter for targeted suppression.

[0134] Frequency connectivity can be a topological characteristic describing whether adjacent frequency points are continuously distributed in the power spectrum. It is determined by analyzing the spatial continuity between frequency points within a frequency band, and can be implemented through sliding window detection or clustering algorithms. For example, frequency connectivity can include adjacency matrix analysis based on the continuity of amplitude of adjacent frequency points or interval merging rules based on frequency interval thresholds. Edge noise region can be a noise frequency band located at the edge of a high-frequency band and exhibiting a dispersed distribution. It may be caused by parasitic parameters of the switching power supply or external electromagnetic interference. For example, edge noise region can include discrete spike noise in the 1MHz to 2MHz frequency band. Harmonic noise region can be a continuous noise frequency band concentrated near a specific harmonic frequency. It may originate from periodic distortion of nonlinear components or integer multiples of the fundamental frequency interference. For example, harmonic noise region can include continuous frequency bands centered at 50kHz and 100kHz. For example, this can be achieved as follows: First, perform a continuous scan on the frequency axis for all frequency bands to be compensated, identify the intervals consisting of continuous frequency points as harmonic noise regions, and classify the remaining dispersed frequency bands as edge noise regions. Then, design an RC lead compensation network for the edge noise regions to improve the high-frequency phase margin. Specifically, adjust the combined parameters of resistors and capacitors to make the compensation network generate phase lead characteristics in the target frequency band. At the same time, design an LC notch filter for the harmonic noise regions to form resonant attenuation. By matching the resonant frequency with the noise center frequency, deep suppression can be achieved, thereby achieving the technical effect of selecting differentiated compensation strategies according to the noise distribution characteristics.

[0135] This embodiment converts the time-domain signal into a frequency-domain power spectrum and locates the frequency band to be compensated based on threshold segmentation. It uses frequency connectivity characteristics to classify the noise frequency band by edge and harmonic type, and adopts an RC lead compensation network to improve high-frequency response stability and an LC notch filter to suppress specific harmonic interference. This achieves the technical effect of accurately locating noise sources and providing targeted compensation. Compared with the traditional global compensation scheme, this method reduces interference to non-noise frequency bands and reduces energy loss through group compensation. At the same time, it improves the accuracy of high-frequency noise suppression and the overall stability of the system through differentiated compensation strategies.

[0136] Furthermore, this embodiment of the invention also proposes a storage medium storing a transient response optimization program based on a linear regulator. When the transient response optimization program based on a linear regulator is executed by a processor, it implements the steps of the transient response optimization method based on a linear regulator as described above.

[0137] In addition, refer to Figure 3 This invention also proposes a transient response optimization system based on a linear regulator, the transient response optimization system based on a linear regulator comprising:

[0138] The early warning judgment module 10 is used to acquire transient response data of the linear regulator when the load changes suddenly, and to determine whether the linear regulator has entered a transient early warning state based on the transient response data;

[0139] The compensation calculation module 20 is used to perform high-frequency noise threshold segmentation on the output voltage waveform of the linear regulator if a transient warning state is entered, extract the voltage change rate characteristics of the threshold segmentation region, construct a compensation efficiency model based on dynamic response time and effective compensation bandwidth through the voltage change rate characteristics, and calculate the optimal open-loop compensation coefficient according to the compensation efficiency model.

[0140] The dynamic compensation module 30 is used to generate compensation control commands based on the optimal open-loop compensation coefficient, and drive the compensation network to perform dynamic compensation on the linear regulator.

[0141] The ripple analysis module 40 is used to acquire the ripple uniformity data of the output voltage of the linear regulator if the transient warning state is not entered, determine the deviation matrix between the actual ripple distribution surface and the preset ideal ripple distribution plane based on the ripple uniformity data, and determine whether there is a local ripple abnormal area based on the deviation matrix.

[0142] The regional compensation module 50 is used to generate a regional compensation adjustment signal if there is a local ripple abnormality area, and control the compensation network to perform directional compensation for the abnormal area until the transient response index of the linear regulator is in the normal operating range, which is a state with no transient warning and uniform ripple.

[0143] Other embodiments or specific implementations of the transient response optimization system based on a linear regulator described in this invention can be found in the above-described method embodiments, and will not be repeated here.

[0144] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or system. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or system that includes that element.

[0145] The sequence numbers of the above embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments. In the unit claims listing several systems, these systems may be embodied by the same hardware item. The use of the terms "first," "second," and "third," etc., does not indicate any order and can be interpreted as names.

[0146] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as a read-only memory image (ROM) / random access memory (RAM), magnetic disk, optical disk), and includes several instructions to cause a terminal user device (which may be a mobile phone, computer, server, air conditioner, or network user device, etc.) to execute the methods described in the various embodiments of the present invention.

[0147] The above are merely preferred embodiments of the present invention and do not limit the scope of the patent. Any equivalent structural or procedural transformations made based on the description and drawings of the present invention, or direct or indirect applications in other related technical fields, are similarly included within the scope of patent protection of the present invention.

Claims

1. A method for optimizing the transient response of a linear voltage regulator, characterized in that, The method includes: Acquire transient response data of the linear regulator when the load changes abruptly, and determine whether the linear regulator has entered a transient warning state based on the transient response data; If a transient warning state is entered, the output voltage waveform of the linear regulator is segmented by a high-frequency noise threshold, and the voltage change rate characteristics of the threshold segmentation region are extracted. A compensation efficiency model based on dynamic response time and effective compensation bandwidth is constructed based on the voltage change rate characteristics, and the optimal open-loop compensation coefficient is calculated based on the compensation efficiency model. Based on the optimal open-loop compensation coefficient, a compensation control command is generated to drive the compensation network to dynamically compensate the linear regulator; If the transient warning state is not entered, the ripple uniformity data of the output voltage of the linear regulator is obtained, and the deviation matrix between the actual ripple distribution surface and the preset ideal ripple distribution plane is determined based on the ripple uniformity data. Based on the deviation matrix, it is determined whether there is a local ripple abnormal area. If there are local ripple anomalies, a regional compensation adjustment signal is generated to control the compensation network to perform directional compensation on the anomaly area until the transient response index of the linear regulator is in the normal operating range. The normal operating range is a state with no transient warning and uniform ripple.

2. The linear regulator-based transient response optimization method of claim 1, wherein, The step of acquiring transient response data of the linear regulator during load changes, and determining whether the linear regulator has entered a transient warning state based on the transient response data, includes: Acquire transient response data of a linear regulator under load change, determine the output voltage waveform data of the linear regulator based on the transient response data, the voltage waveform data includes a time-domain voltage sequence, and extract edge voltage sampling point information of rising and falling edges in the time-domain voltage sequence; The voltage change rate of adjacent sampling points at the rising and falling edges is calculated based on the edge voltage sampling point information, and a slope stability index is constructed based on the voltage change rate of the rising and falling edges. The slope stability index is used as a transient response warning criterion. The difference between the transient response warning criterion and the preset transient threshold is used to determine whether the linear regulator has entered a transient warning state.

3. The linear regulator-based transient response optimization method of claim 2, wherein, The slope stability index constructed based on the voltage change rate at the rising and falling edges includes: The magnitude and direction of the dynamic response gradient are determined based on the rate of change of voltage at the rising and falling edges. The sampling points are grouped according to the gradient direction, and the mean gradient magnitude difference and the mean direction deviation angle of adjacent sampling points in each group are calculated. The slope stability index is calculated based on the mean gradient magnitude difference and the mean direction deviation angle, wherein the formula for calculating the stability index is: In the formula, S is a stability index, the larger the value, the more stable the slope, ΔG is the gradient amplitude difference value average, is the average of the directional deviation angle.

4. The linear regulator-based transient response optimization method of claim 1, wherein, The process involves constructing a compensation efficiency model based on dynamic response time and effective compensation bandwidth using the voltage change rate characteristics, and calculating the optimal open-loop compensation coefficient based on the compensation efficiency model, including: Based on the voltage change rate characteristics, frequency domain analysis is performed on the compensation region after threshold segmentation to calculate the dynamic response time and effective compensation bandwidth. A compensation factor is then defined based on the dynamic response time and effective compensation bandwidth, with the following formula: wherein is a compensation factor, is a regulator nominal bandwidth, is a dynamic response time, is an effective compensation bandwidth; establishing a compensation efficiency model between the compensation factor and the optimal open-loop compensation coefficient, the compensation efficiency model including a mapping relationship between the compensation factor and the optimal open-loop compensation coefficient, and calculating the optimal open-loop compensation coefficient satisfying a target response accuracy according to the mapping relationship; wherein a calculation formula of the mapping relationship is: In the formula, is the optimal open-loop compensation coefficient, is the initial compensation coefficient, is the compensation sensitivity coefficient, determined by the regulator load characteristic and output capacitance parameter, is the compensation factor.

5. The linear regulator-based transient response optimization method of claim 1, wherein, determining a deviation matrix of an actual ripple distribution surface and a preset ideal ripple distribution plane based on the ripple uniformity data, and judging whether there is a local ripple abnormal area according to the deviation matrix, including: converting the ripple uniformity data into a frequency domain ripple spectrum matrix, and fitting to obtain an actual ripple distribution surface; performing frequency domain registration on the actual ripple distribution surface and the preset ideal ripple plane, and calculating a ripple amplitude deviation of each frequency point; constructing a deviation matrix based on the ripple amplitude deviation, and determining a frequency point proportion and energy distribution whose deviation absolute value exceeds a preset ripple threshold value; judging whether there is a local ripple abnormal area according to the frequency point proportion and the energy distribution.

6. The linear regulator-based transient response optimization method of claim 1, wherein, if there is a local ripple abnormal area, generating a regional compensation adjustment signal, and controlling the compensation network to perform directional compensation on the abnormal area, including: if there is a local ripple abnormal area, generating a frequency weight matrix according to the deviation matrix, and the weight value is positively correlated with the ripple amplitude deviation absolute value; generating a regional compensation adjustment signal based on the frequency weight matrix, adjusting the band-pass filter parameter of the compensation network, and the adjustment intensity is proportional to the gradient value of the weight matrix, and the directional compensation is preferentially performed on the high-weight abnormal frequency segment.

7. The linear regulator-based transient response optimization method of claim 1, wherein, generating a compensation control instruction according to the optimal open-loop compensation coefficient, and driving the compensation network to perform dynamic compensation on the linear voltage regulator, including: generating a compensation control instruction according to the optimal open-loop compensation coefficient, and real-time monitoring of input voltage fluctuation and load current change, and dynamically adjusting the compensation energy distribution priority; driving the adjustable resistance-capacitance array in the compensation network based on the compensation control instruction and the energy distribution priority, to realize dynamic compensation on the loop gain of the linear voltage regulator.

8. The linear regulator-based transient response optimization method of claim 7, wherein, dynamically adjusting the compensation energy distribution priority, including: when the input voltage fluctuation amplitude exceeds the preset voltage amplitude, a high-frequency compensation path is preferentially enabled to suppress voltage overshoot, and the high-frequency compensation path includes an RC parallel network; when the load current change rate exceeds the preset current change rate threshold, a low-frequency compensation path is switched to optimize the steady-state error, and the low-frequency compensation path includes an LC series network.

9. The linear regulator-based transient response optimization method of claim 1, wherein, performing high-frequency noise threshold segmentation on the linear voltage regulator output voltage waveform, including: converting the linear voltage regulator output voltage waveform into a frequency domain power spectrum, setting a high-frequency noise warning threshold as a segmentation reference, and marking a frequency band exceeding the reference as a to-be-compensated frequency band; grouping the to-be-compensated frequency band according to frequency connectivity, including an edge noise zone and a harmonic noise zone, the edge noise zone adopts an RC lead compensation network, and the harmonic noise zone adopts an LC notch filter for targeted suppression.

10. A system for optimizing transient response based on a linear voltage regulator, the system comprising: the transient response optimization system based on the linear voltage regulator includes: The early warning judgment module is configured to acquire transient response data of the linear voltage stabilizer when a load is suddenly changed, and determine whether the linear voltage stabilizer enters a transient early warning state based on the transient response data. The compensation calculation module is configured to, if the linear voltage stabilizer enters the transient early warning state, perform high-frequency noise threshold segmentation on an output voltage waveform of the linear voltage stabilizer, extract a voltage change rate feature of a threshold segmentation region, construct a compensation efficiency model based on a dynamic response time and an effective compensation bandwidth through the voltage change rate feature, and calculate an optimal open-loop compensation coefficient according to the compensation efficiency model. The dynamic compensation module is configured to generate a compensation control instruction according to the optimal open-loop compensation coefficient, and drive a compensation network to perform dynamic compensation on the linear voltage stabilizer. The ripple analysis module is configured to, if the linear voltage stabilizer does not enter the transient early warning state, acquire ripple uniformity data of an output voltage of the linear voltage stabilizer, determine a deviation matrix of an actual ripple distribution surface and a preset ideal ripple distribution plane based on the ripple uniformity data, and determine whether there is a local ripple abnormal region according to the deviation matrix. The region compensation module is configured to, if there is a local ripple abnormal region, generate a region compensation adjustment signal, control the compensation network to perform directional compensation on the abnormal region, and keep the transient response index of the linear voltage stabilizer in a normal working interval until the linear voltage stabilizer is in a state of no transient early warning and uniform ripple.

Images