A seamless switching control method for multi-mode operation state of a UPS system

Abstract

The application discloses a seamless switching control method for multi-mode operation state of a UPS system and belongs to the technical field of UPS power supply control. The method acquires a real-time drive duty ratio sequence of a power switch tube and a DC bus voltage ripple sequence in a current operation mode, constructs a duty ratio disturbance intensity and a ripple fluctuation index, inputs the duty ratio disturbance intensity and the ripple fluctuation index into a hysteresis comparator, and generates a mode switching pre-trigger flag. When the flag is valid, a phase shift offset of an inverter voltage zero-crossing point is extracted to correct a phase output of a phase-locked loop, a transition drive duty ratio is generated, and then the transition drive duty ratio is weighted and averaged with a target steady-state drive duty ratio. A weighted coefficient is dynamically determined by a ratio of the disturbance intensity to the ripple fluctuation index, and a seamless switching control signal is obtained. The method can actively perceive a state change trend before mode switching, suppress phase jump and control quantity mutation in the switching process, and realize smooth seamless switching.

Description

Technical Field

[0001] This invention relates to the field of UPS power control technology, specifically to a seamless switching control method for multiple operating modes of a UPS system. Background Technology

[0002] UPS power systems need to switch between multiple operating modes, including normal mains power, abnormal mains power, battery discharge, and bypass, to ensure the continuity of power supply to the load and power quality. Maintaining the continuity of the inverter output voltage waveform and avoiding voltage dips or phase jumps caused by abrupt changes in control mode during multi-mode switching has long been a technical challenge in this field. Most existing multi-mode switching control methods use threshold comparisons based on voltage amplitude or frequency for triggering. When the output voltage or frequency in the current mode exceeds a preset range, the system performs a mode switch. This type of method is a passive response; the switching action is only triggered after a voltage anomaly has occurred, resulting in a delayed switching timing. Furthermore, at the moment of switching, the drive signal often abruptly changes from the open-loop or closed-loop control quantity of the current mode to the initial control quantity of the target mode, causing a step in duty cycle and phase, resulting in distortion or even interruption of the output voltage waveform, impacting sensitive loads.

[0003] Some improved solutions attempt to reduce phase discontinuities by continuously tracking the grid phase using a phase-locked loop (PLL). However, at the moment the mode switching command takes effect, a static deviation may still exist between the PLL output phase and the actual inverter voltage phase, and direct switching will still introduce phase disturbances. Other solutions employ a ramp function to gradually adjust the reference voltage or current, allowing the control quantity to transition smoothly over a certain period. However, such open-loop gradual adjustment strategies lack closed-loop perception of dynamic changes in the actual operating state, and new disturbances may be superimposed during the transition, affecting the smoothness of the switching.

[0004] Therefore, the key problem to be solved is how to quickly identify the precursory characteristics of the current operating mode that are about to become unstable before the mode switch occurs, and on this basis, comprehensively handle the phase synchronization and control quantity connection to achieve smooth closed-loop control from switching prediction to the transition process. Summary of the Invention

[0005] This invention provides a seamless switching control method for multi-mode operation of a UPS system. The purpose is to detect the evolution trend of the system state in advance and actively generate a switching pre-trigger flag in multi-mode switching scenarios. At the same time, through phase correction and dynamic weighted fusion, the phase discontinuity and duty cycle change during the switching process are eliminated, so as to achieve a smooth and seamless transition of the output voltage waveform.

[0006] To achieve the above objectives, the present invention provides the following technical solution: The present invention provides a seamless switching control method for multi-mode operation of a UPS system, comprising the following steps: acquiring the real-time drive duty cycle sequence and DC bus voltage ripple sequence of each power switch in the current operating mode of the UPS system. Constructing a duty cycle disturbance intensity based on the frequency of sign changes in the difference between adjacent duty cycles in the real-time drive duty cycle sequence. Constructing a ripple fluctuation index based on the dispersion of the ratio sequence of adjacent ripple amplitudes in the DC bus voltage ripple sequence. Inputting the duty cycle disturbance intensity and the ripple fluctuation index into a hysteresis comparator to generate a mode switching pre-trigger flag. When the mode switching pre-trigger flag is valid, extracting the phase offset of the inverter voltage zero-crossing point within one power frequency cycle prior to the current moment. Correcting the output phase of the phase-locked loop based on the phase offset, and generating the transition drive duty cycle of each power switch based on the corrected output phase. The transition drive duty cycle is weighted and averaged with the steady-state drive duty cycle of the target operating mode. The weighting coefficient is dynamically determined by the ratio of the duty cycle disturbance intensity and the ripple fluctuation index to obtain a seamless switching control signal.

[0007] When constructing the duty cycle perturbation strength, the real-time drive duty cycle sequence is arranged in ascending order of sampling time. The difference between each subsequent duty cycle and the previous duty cycle is calculated sequentially to obtain a difference sequence. The number of times the product of two adjacent differences in the difference sequence is negative is recorded as the sign inversion count. The ratio of the sign inversion count to the total number of non-zero differences in the difference sequence is determined as the duty cycle perturbation strength. This method captures the subtle oscillation tendency of the drive signal, providing a sensitive perturbation indicator for mode switching.

[0008] When constructing the ripple fluctuation index, the ripple amplitudes in the DC bus voltage ripple sequence are arranged in chronological order, and the ratio of each subsequent ripple amplitude to the previous ripple amplitude is calculated sequentially to obtain a ratio sequence. The standard deviation of all ratios in the ratio sequence is then calculated to obtain the ratio standard deviation. The ratio of the ratio standard deviation to the mean of all ratios in the ratio sequence is determined as the ripple fluctuation index. Preferably, the absolute values ​​of each ratio in the ratio sequence are taken before calculating the standard deviation and mean to eliminate interference from ripple polarity changes and to truly reflect the degree of power quality fluctuation.

[0009] The process of generating the mode switching pre-trigger flag is as follows: Calculate the difference between the duty cycle disturbance intensity and a first threshold to obtain a first deviation; calculate the difference between the ripple fluctuation index and a second threshold to obtain a second deviation; input the first and second deviations into two independent hysteresis comparators respectively to obtain a first hysteresis output value and a second hysteresis output value; when both the first and second hysteresis output values ​​are high, the mode switching pre-trigger flag is set to an active state. Through dual hysteresis comparison, false triggering caused by instantaneous disturbances of a single indicator can be effectively avoided, improving the reliability of the switching determination. Preferably, the two independent hysteresis comparators have the same hysteresis width, and when either hysteresis output value transitions from high to low, the mode switching pre-trigger flag is synchronously reset, ensuring that switching preparation is only initiated when both disturbance indicators continuously exceed the threshold.

[0010] The steps for extracting the phase offset include: recording the actual zero-crossing time of each inverter voltage zero-crossing point within one power frequency cycle prior to the current moment; obtaining the time difference between each actual zero-crossing time and the corresponding ideal zero-crossing time to obtain a zero-crossing time difference sequence; dividing each zero-crossing time difference in the zero-crossing time difference sequence by the power frequency cycle value and multiplying by 360 degrees to obtain the instantaneous phase offset of each zero-crossing point; and determining the average of all instantaneous phase offsets as the phase offset. The multi-zero-crossing averaging method can filter out random noise and improve the accuracy of phase offset estimation.

[0011] When correcting the output phase of the phase-locked loop (PLL) based on the phase offset and generating the transition drive duty cycle, the current output phase of the PLL is added to the phase offset to obtain the compensated reference phase. The switching angle sequence of each power switch is then obtained from a table based on the compensated reference phase. This switching angle sequence is converted into the duty cycle of each power switch in the next switching cycle to obtain the transition drive duty cycle. This process quickly corrects the PLL output, ensuring phase continuity at the start of the switching.

[0012] In weighted averaging, the ratio of duty cycle disturbance intensity to ripple fluctuation index is calculated to obtain the dynamic ratio. This dynamic ratio is then mapped to a closed interval of zero to one to obtain the weighting coefficient. The weighting coefficient is multiplied by the transition drive duty cycle, and the difference between one and the weighting coefficient is multiplied by the steady-state drive duty cycle. The products are then added together to obtain the seamless switching control signal. Preferably, the dynamic ratio is mapped to a closed interval of zero to one through a saturation function, and when the dynamic ratio exceeds a preset upper limit, the weighting coefficient is directly set to one. The transition degree is dynamically adjusted according to the relative strength of the disturbance and ripple. When the disturbance is severe, a rapid transition duty cycle is used to achieve a fast and smooth switching. When the disturbance is mild, the steady-state duty cycle is gradually integrated to suppress voltage and current surges, ensuring a smooth and seamless switching process.

[0013] The technical effects and advantages provided by the present invention in the above technical solution are as follows:

[0014] By converting the frequency of sign changes of adjacent duty cycle differences in the real-time drive duty cycle sequence into duty cycle disturbance intensity, and by converting the dispersion of the ratio of adjacent ripple amplitudes in the DC bus voltage ripple sequence into a ripple fluctuation index, sensitive features reflecting the degradation of the stability margin of the current operating mode can be extracted from both the switching drive signal and the bus voltage ripple. Compared to macroscopic electrical quantities such as voltage amplitude and frequency, these two features have a higher predictive ability for micro-oscillations and instability trends in the control loop within the mode. Inputting the duty cycle disturbance intensity and ripple fluctuation index into a hysteresis comparator generates a mode switching pre-trigger flag, ensuring that the switching decision does not wait for voltage limits to exceed, but rather is proactively triggered when the system's internal dynamic characteristics begin to deteriorate, providing an advanced time window for seamless switching. This pre-trigger mechanism based on dual-dimensional disturbance characteristics and hysteresis judgment effectively avoids voltage drops or waveform distortion caused by switching lag, improving the predictability of mode switching and power supply continuity. After the mode switching pre-trigger flag is valid, the phase offset of the inverter voltage zero-crossing sequence in the previous power frequency cycle is extracted and directly corrected to the output phase of the phase-locked loop (PLL), aligning the compensated reference phase with the true phase of the current inverter voltage. The transition drive duty cycle of each power switch is generated by looking up the corrected phase in a table. This method eliminates phase angle jumps caused by ideal PLL tracking deviations at the source. Furthermore, the transition drive duty cycle is weighted and averaged with the steady-state drive duty cycle of the target operating mode. The weighting coefficient is dynamically determined by the ratio of the duty cycle disturbance intensity to the ripple fluctuation index. When the disturbance is severe, the dynamic ratio increases, and the weighting coefficient is more biased towards the transition drive duty cycle to prioritize the continuity of the inverter output. As the disturbance gradually converges, the weighting coefficient smoothly shifts towards the steady-state drive duty cycle, allowing the control quantity to naturally transition to the stable operating point of the target mode. This fusion strategy achieves continuous change of duty cycle throughout the switching process, avoiding output shock caused by sudden changes in control quantity. The output waveform does not produce perceptible discontinuities in amplitude, phase and frequency, achieving truly seamless switching between multiple operating modes. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0016] Figure 1 It is a flowchart of the seamless switching control method for multiple operating modes of a UPS system;

[0017] Figure 2 This is a flowchart of the duty cycle disturbance intensity calculation method;

[0018] Figure 3 This is a schematic diagram of the real-time driven duty cycle sequence and sign inversion point;

[0019] Figure 4 It is a curve showing the change in DC bus voltage ripple amplitude and ripple fluctuation index. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] See Figure 1 This invention provides a seamless switching control method for multi-mode operation of a UPS system, comprising: acquiring the real-time drive duty cycle sequence and DC bus voltage ripple sequence of each power switch in the current operating mode of the UPS system; constructing the duty cycle disturbance intensity based on the frequency of sign change of the difference between adjacent duty cycles in the real-time drive duty cycle sequence; constructing the ripple fluctuation index based on the dispersion of the ratio sequence of adjacent ripple amplitudes in the DC bus voltage ripple sequence; inputting the duty cycle disturbance intensity and the ripple fluctuation index into a hysteresis comparator to generate a mode switching pre-trigger flag; when the mode switching pre-trigger flag is valid, extracting the phase offset of the inverter voltage zero-crossing point in the previous power frequency cycle; correcting the output phase of the phase-locked loop based on the phase offset, and generating the transition drive duty cycle of each power switch based on the corrected output phase; weighting the transition drive duty cycle with the steady-state drive duty cycle of the target operating mode, wherein the weighting coefficient is dynamically determined by the ratio of the duty cycle disturbance intensity and the ripple fluctuation index to obtain a seamless switching control signal.

[0022] In specific implementation, please refer to Figure 2 This method acquires the real-time drive duty cycle sequence of each power switch in the UPS system under the current operating mode. The real-time drive duty cycle sequence consists of duty cycle values ​​obtained from consecutive sampling times, with each duty cycle value corresponding to one sampling time.

[0023] The real-time drive duty cycle sequence is arranged in ascending order of sampling time to obtain an ascending real-time drive duty cycle sequence. Within this ascending sequence, the difference between each subsequent duty cycle and the previous one is calculated. Let the ascending real-time drive duty cycle sequence be... ,in, Let represent the real-time drive duty cycle at the i-th sampling time, and n represent the total number of duty cycles in the real-time drive duty cycle sequence. Calculate... and The difference and The difference, until the calculation and The difference is represented by a sequence of all the differences. Each element in the difference sequence is the difference between two adjacent duty cycles.

[0024] The number of times the product of two adjacent differences in the difference sequence is negative is denoted as the sign inversion count. After obtaining the difference sequence, multiply every two adjacent differences. When the product is negative, it indicates that the signs of the two differences are opposite, meaning the direction of the duty cycle change has reversed once. The total number of negative product results is the sign inversion count. Simultaneously, identify all non-zero differences in the difference sequence and count them to obtain the total number of non-zero differences. The ratio of the sign inversion count to the total number of non-zero differences is determined as the duty cycle perturbation strength.

[0025] Optional, use The duty cycle disturbance strength is expressed by the following formula:

[0026]

[0027] in, This indicates the number of sign reversals, which is the statistical number of times the product of two adjacent differences in the difference sequence is negative. This represents the total number of non-zero differences, which is the statistical count of all non-zero differences in the difference sequence.

[0028] In some embodiments, when the difference between adjacent duty cycles in the real-time drive duty cycle sequence is all zero, the total number of non-zero differences is zero, and the duty cycle disturbance intensity is directly set to zero.

[0029] It is understandable that an increase in the number of sign inversions implies frequent direction switching of the duty cycle sequence within a short period, reflecting the instability of the driving signal. The duty cycle perturbation strength, as a scalar indicator, combines the frequency of sign inversions with the proportion of non-zero changes.

[0030] See Figure 3In the figure, the horizontal axis represents the sampling sequence number, and the vertical axis represents the duty cycle and the percentage value of the difference, reflecting the variation characteristics of the real-time drive duty cycle sequence and the difference between adjacent duty cycles obtained in the above embodiment. The black solid line represents the real-time drive duty cycle, which fluctuates between approximately 45% and 55%, showing a periodic and small-amplitude oscillation trend, indicating that the drive duty cycle of the power switch remains relatively stable during the sampling time but has minor disturbances. The blue dashed line represents the sequence of differences between adjacent duty cycles, which fluctuates around zero with small amplitude, indicating that the change in duty cycle is relatively gentle. The red crosses mark the sign reversal points, which are scattered at the boundaries of positive and negative changes in adjacent duty cycle difference sequences. Their number is moderate and their distribution is relatively even, reflecting the statistical characteristics of the number of sign reversals.

[0031] In practical implementation, the DC bus voltage ripple sequence of the UPS system under the current operating mode is obtained. The DC bus voltage ripple sequence consists of ripple amplitude values ​​obtained at consecutive sampling times, with each ripple amplitude value corresponding to one sampling time.

[0032] Arranging the ripple amplitudes in the DC bus voltage ripple sequence in chronological order yields the time-sequential DC bus voltage ripple sequence. Let the time-sequential DC bus voltage ripple sequence be... ,in, Let represent the DC bus voltage ripple amplitude at the u-th sampling time, and m represent the total number of ripple amplitudes in the DC bus voltage ripple sequence. The ratio of each subsequent ripple amplitude to the previous ripple amplitude is calculated sequentially, i.e., the ratio of each subsequent ripple amplitude to the previous ripple amplitude is calculated. and The ratio, and The ratio until the calculation and The ratios are formed by all the ratios, which constitute a ratio sequence.

[0033] The ratio sequence is formed by sequentially arranging the ratios of two adjacent ripple amplitudes. The absolute value of each ratio in the obtained ratio sequence is taken to obtain an absolute value ratio sequence. Based on the absolute value ratio sequence, the standard deviation of all ratios in the absolute value ratio sequence is obtained as the ratio standard deviation, and the mean of all ratios in the absolute value ratio sequence is obtained as the ratio mean.

[0034] The ratio of the standard deviation to the mean of the ratio is defined as the ripple fluctuation index, which characterizes the dispersion of the ratio of adjacent ripple amplitudes in the DC bus voltage ripple sequence. Optionally, [the index can be used...] The ripple fluctuation index is expressed as follows:

[0035]

[0036] in, The standard deviation of the ratio is the standard deviation of all ratios in the absolute value ratio sequence. The ratio mean is the average of all ratios in the absolute value ratio sequence. The ratio standard deviation is calculated using the standard deviation formula, which is the arithmetic square root of the sum of the squares of the differences between each ratio in the absolute value ratio sequence and the ratio mean. The ratio mean is calculated using the arithmetic mean method, which is obtained by dividing the sum of each ratio in the absolute value ratio sequence by the total number of ratios in the absolute value ratio sequence.

[0037] In some embodiments, when the ratios of adjacent ripple amplitudes in the ratio sequence change drastically, the dispersion of the ratios in the absolute value ratio sequence increases, and the standard deviation of the ratios increases accordingly. When the mean of the ratios is not zero, the ripple fluctuation index increases accordingly. When there is only one ratio or all ratios are equal in the absolute value ratio sequence, the standard deviation of the ratios is zero, and the ripple fluctuation index is zero.

[0038] It can be understood that each ratio in the ratio sequence reflects the relative change in the DC bus voltage ripple amplitude between adjacent sampling times. By taking the absolute value of the ratio, the influence of the ripple change direction on the dispersion analysis is eliminated, making the statistical characteristics of the absolute value ratio sequence only related to the relative magnitude of the ripple amplitude change.

[0039] See Figure 4 In the figure, the horizontal axis represents the sampling sequence number, indicating the order of continuous sampling times, with a value range of approximately 0 to 800. The left vertical axis represents the DC bus voltage ripple amplitude, in volts (V), and the right vertical axis represents the ripple fluctuation index and the absolute value of the ratio of adjacent ripple amplitudes, in ratio / ripple fluctuation index. The black solid line in the figure represents the trend of the DC bus voltage ripple amplitude with the sampling sequence number, exhibiting periodic fluctuation characteristics, with the amplitude fluctuating in the range of approximately -1.6V to 1.8V, reflecting the time-domain characteristics of the DC bus voltage ripple. The blue dashed line represents the absolute value of the ratio of adjacent ripple amplitudes, which remains at a low and stable level overall, with occasional short-term peak surges, indicating that the ripple amplitude changes smoothly most of the time, with occasional significant fluctuations. The red dotted line represents the ripple fluctuation index, which remains at a low level close to zero, indicating that the dispersion of the ripple amplitude ratio sequence is small and the ripple stability is good.

[0040] Based on the specific technical solution of the above embodiment, this figure reflects the dispersion of ripple amplitude changes and the dynamic characteristics of ripple fluctuation index by calculating and statistically analyzing the ratio of adjacent ripple amplitudes in the DC bus voltage ripple sequence. The periodic fluctuations of the black ripple amplitude curve in the figure correspond to the occasional peaks of the blue adjacent ripple amplitude ratio curve, indicating that at sampling numbers approximately 150, 250, and 700, the ripple amplitude undergoes relatively drastic changes, leading to a short-term significant increase in the absolute value of the ratio of adjacent ripple amplitudes. However, the overall ripple fluctuation index does not show a significant increase, demonstrating that the ripple fluctuation index calculation method used in this embodiment has a certain robustness and suppression effect on short-term drastic fluctuations.

[0041] In practical implementation, the duty cycle disturbance intensity and ripple fluctuation index are obtained. The duty cycle disturbance intensity is constructed by the frequency of sign changes in the difference between adjacent duty cycles in the real-time drive duty cycle sequence. The ripple fluctuation index is constructed by the dispersion of the ratio sequence of adjacent ripple amplitudes in the DC bus voltage ripple sequence.

[0042] Set a first threshold and a second threshold. The first threshold is a predetermined limit value for duty cycle disturbance intensity, used for comparison with the duty cycle disturbance intensity; the second threshold is a predetermined limit value for ripple fluctuation index, used for comparison with the ripple fluctuation index. The values ​​of the first threshold and the second threshold are determined based on the rated parameters of the UPS system and the allowable disturbance tolerance.

[0043] The difference between the duty cycle disturbance intensity and the first threshold is calculated to obtain the first deviation. The first deviation is represented by the duty cycle disturbance strength. In other words, the first threshold is represented by... This indicates that the first deviation is obtained in the following way: The difference between the ripple fluctuation index and the second threshold is calculated to obtain the second deviation. The second deviation is represented by the ripple fluctuation index. This indicates that the second threshold is used This indicates that the second deviation is obtained in the following way: .

[0044] The first and second deviations are input to two independent hysteresis comparators, respectively. In each comparator, the first comparator receives the first deviation as its input signal, and the second comparator receives the second deviation as its input signal. Both comparators have the same hysteresis width, which is represented by the formula... This indicates that the hysteresis width is set according to the allowable deviation fluctuation range. In the first hysteresis comparator, when the first deviation increases by more than... When the first hysteresis output value is set to high, the first deviation value is set to low. When the first hysteresis output value is set to low, the first deviation value is set to low. to Within the specified interval, the first hysteresis output value remains unchanged from its previous state. The second hysteresis comparator processes the second deviation in the same way as the first hysteresis comparator processes the first deviation, wherein the second hysteresis output value is determined by the second deviation and the hysteresis width. The comparison results determine the outcome.

[0045] The system acquires the first hysteresis output value and the second hysteresis output value. When both the first and second hysteresis output values ​​are high, the mode switching pre-trigger flag is set to an active state. The active state of the mode switching pre-trigger flag indicates that the current operating state of the UPS system meets the pre-trigger conditions for mode switching. When either the first or second hysteresis output value transitions from high to low, the mode switching pre-trigger flag is synchronously reset. Synchronous reset means that the mode switching pre-trigger flag is set to an inactive state within the same control cycle during which either hysteresis output value transitions from high to low.

[0046] In some embodiments, the mode switching pre-trigger flag is generated by an AND gate circuit, the first hysteresis output value and the second hysteresis output value are respectively connected to the two input terminals of the AND gate circuit, and the output terminal of the AND gate circuit outputs the mode switching pre-trigger flag.

[0047] It is understandable that the hysteresis width of the hysteresis comparator... The setting avoids frequent flipping of the first hysteresis output value or the second hysteresis output value when the first deviation or the second deviation fluctuates slightly near the corresponding threshold, thereby preventing the mode switching pre-trigger flag from repeatedly jumping between the valid state and the invalid state.

[0048] In practical implementation, when the mode switching pre-trigger flag is valid, the phase offset of the inverter voltage zero-crossing point within the previous power frequency cycle is extracted. The inverter voltage zero-crossing point refers to the moment when the inverter output voltage waveform crosses zero voltage as it changes from the negative half-cycle to the positive half-cycle. The ideal zero-crossing time refers to the reference time when the inverter voltage waveform should cross zero voltage under ideal grid synchronization conditions. The ideal zero-crossing time is obtained recursively from the nominal power frequency cycle and the reference phase.

[0049] Record the actual zero-crossing time of each inverter voltage zero-crossing point within one power frequency cycle prior to the current moment. The zero-crossing detection circuit captures the instant the inverter voltage waveform crosses zero voltage, and records this capture moment as the actual zero-crossing time. Within a complete power frequency cycle, there are two zero-crossing points for a single-phase system and two zero-crossing points for each phase of a three-phase system. All recorded actual zero-crossing times are arranged in chronological order of occurrence.

[0050] Obtain the time difference between each actual zero-crossing moment and its corresponding ideal zero-crossing moment to obtain a zero-crossing time difference sequence. Let the total number of recorded inverter voltage zero-crossing points be... , No. The actual zero-crossing time corresponding to each inverter voltage zero-crossing point is: , No. The ideal zero-crossing time corresponding to the zero-crossing point of each inverter voltage is: Then the first The time difference of the inverter voltage zero crossing point The calculation method is as follows The sequence consisting of the time differences of all zero-crossing points is called the zero-crossing time difference sequence, denoted as: .

[0051] Divide each zero-crossing time difference in the zero-crossing time difference sequence by the period of the power frequency cycle and multiply by 360 degrees to obtain the instantaneous phase shift at each zero-crossing point. (Using the period of the power frequency cycle...) This represents the time length corresponding to a complete power frequency cycle, in seconds. Instantaneous phase shift at zero crossing point Expressed in electrical degrees, its calculation expression is: The average of all instantaneous phase shifts is defined as the phase shift. The phase shift is expressed as... It indicates that the method of acquisition is ,in express From 1 to The sum of all instantaneous phase shifts.

[0052] The current output phase of the phase-locked loop (PLL) is added to the phase offset to obtain the compensated reference phase. The current output phase of the PLL is the phase value output by the PLL at the moment the phase offset is acquired. The compensated reference phase is represented by... It indicates that the method of acquisition is .

[0053] The switching angle sequence of each power switch is obtained by looking up a table based on the compensated reference phase. The switching angle sequence is stored in a pre-calculated lookup table, which takes the reference phase as input and outputs the turn-on and turn-off angles of each power switch within one power frequency cycle. For each power switch, the lookup table outputs its corresponding set of turn-on angles and set of turn-off angles, forming the switching angle sequence.

[0054] The switching angle sequence is converted into the duty cycle of each power switch in the next switching cycle to obtain the transition drive duty cycle. The switching cycle is the pulse width modulation carrier cycle of the power switch. For each power switch, based on the turn-on and turn-off angles corresponding to the next switching cycle in the switching angle sequence, the proportion of the power switch that should be in the on state within that switching cycle is calculated. This proportion is the transition drive duty cycle of the corresponding power switch.

[0055] In some embodiments, the lookup table is constructed as follows: the optimal switching angle of each power switch corresponding to different reference phases under ideal power grid conditions is calculated offline, and the reference phase is discretized with a preset step size. Each discretized reference phase value corresponds to a set of switching angle values, and all discretized reference phase values ​​and their corresponding switching angle values ​​together constitute the lookup table. During online table lookup, the compensated reference phase is mapped to the existing discrete reference phase values ​​in the lookup table according to the nearest neighbor principle or linear interpolation principle, thereby indexing the corresponding switching angle sequence.

[0056] It is understandable that obtaining the average instantaneous phase shift through the zero-crossing time difference sequence can smooth out the random errors that may exist in the detection of a single zero-crossing point, making the phase shift more accurately reflect the overall phase deviation between the inverter voltage and the ideal grid voltage. After the current output phase of the phase-locked loop is compensated for the phase shift, the compensated reference phase eliminates the phase deviation, enabling the pulse width modulation waveform generated by the transition drive duty cycle to maintain continuity with the grid voltage phase of the target operating mode.

[0057] In practical implementation, the duty cycle disturbance intensity and ripple fluctuation index are obtained. The duty cycle disturbance intensity is constructed by the frequency of sign changes in the difference between adjacent duty cycles in the real-time drive duty cycle sequence. The ripple fluctuation index is constructed by the dispersion of the ratio sequence of adjacent ripple amplitudes in the DC bus voltage ripple sequence.

[0058] Calculate the ratio of duty cycle disturbance intensity to ripple fluctuation index to obtain the dynamic ratio. The dynamic ratio, duty cycle, and disturbance intensity are expressed using... The ripple fluctuation index is used to indicate that the ... The expression for calculating the dynamic ratio is:

[0059]

[0060] in, It is a dynamic ratio, a non-negative real value, reflecting the proportional relationship between the duty cycle disturbance intensity and the ripple fluctuation index; The duty cycle perturbation strength is determined by the ratio of the number of sign inversions to the total number of non-zero differences in the difference sequence; The ripple volatility index is determined by the ratio of the standard deviation of the ratios to the mean of the ratios in the absolute value ratio series. When the ripple volatility index... When it is zero, the dynamic ratio We directly take the positive infinity marker value, and the corresponding weighting coefficient for the positive infinity marker value is one in the subsequent mapping process.

[0061] The dynamic ratio is mapped to a closed interval between zero and one using a saturation function to obtain the weighting coefficients. The upper limit of the saturation function is set to one, and the lower limit is set to zero. The input to the saturation function is the dynamic ratio. The output is a weighted coefficient. The mapping rule for the saturation function is: when the dynamic ratio When the weighting coefficient is less than or equal to zero, The value is zero; when the dynamic ratio When the weighting coefficient is greater than zero and less than the preset upper limit, The value and dynamic ratio The relationship is linear, and the proportional coefficient of the linear relationship is determined based on the reciprocal of a preset upper limit; when the dynamic ratio When the weighting coefficient is greater than or equal to the preset upper limit, The value is directly set to 1. The preset upper limit is a pre-set dynamic ratio limit value. The preset upper limit is set based on the fact that when the ratio of the duty cycle disturbance intensity to the ripple fluctuation index exceeds a certain value under the rated operating conditions of the UPS system, it indicates that the reliability of the transition drive duty cycle is high enough, and at this time the transition drive duty cycle is used as the control basis.

[0062] In some embodiments, weighting coefficients Ratio with dynamics The mapping relationship is expressed in piecewise function form: when hour, ;when hour, ,in Indicates the preset upper limit; when hour, Preset upper limit The value range is determined based on the switching frequency of the power switching transistors in the UPS system and the capacitance value of the DC bus capacitor. For applications with a switching frequency in the range of 10kHz to 20kHz and a DC bus capacitor capacitance value in the range of 470μF to 2200μF, a preset upper limit is provided. The setting is 3 because when the dynamic ratio exceeds 3 within this range, the duty cycle disturbance intensity has become significantly dominant, and the influence of ripple fluctuations is secondary.

[0063] Multiplying the weighting coefficients by the transition drive duty cycle yields the first product term. The transition drive duty cycle is generated based on the corrected output phase and corresponds to the drive duty cycle of each power switch. Multiplying the difference (subtracting the weighting coefficients) by the steady-state drive duty cycle yields the second product term. The steady-state drive duty cycle is the pre-calculated drive duty cycle of each power switch under ideal grid synchronization conditions in the target operating mode. Adding the first and second product terms yields the seamless switching control signal. The seamless switching control signal directly acts on the drive circuit of each power switch, controlling the on and off states of each power switch during the switching process.

[0064] Optional, use This indicates seamless switching of control signals, with transition drive duty cycle used... This indicates that the steady-state drive duty cycle is used... Indicates that the weighting coefficients are used This indicates that the seamless switching control signal is generated in the following way: ,in, When the weighting coefficients When the weighting coefficient is zero, the seamless switching control signal is entirely determined by the steady-state drive duty cycle; when the weighting coefficient is zero, the seamless switching control signal is determined entirely by the steady-state drive duty cycle. When the value is 1 minus zero, the seamless switching control signal is entirely determined by the transition drive duty cycle; when the weighting coefficient... When the open interval is between zero and one, the seamless switching control signal is a weighted fusion result of the transition drive duty cycle and the steady-state drive duty cycle.

[0065] In some embodiments, weighting coefficients Updated in real time during each switching cycle. Dynamic ratio. The weighting coefficients vary with the real-time changes in duty cycle disturbance intensity and ripple fluctuation index. Then, a complete calculation and update process is completed within one switching cycle.

[0066] It is understandable that the weighting coefficient is dynamically determined by the ratio of the duty cycle disturbance intensity to the ripple fluctuation index. This allows the transition drive duty cycle to have a greater weight in the seamless switching control signal when the disturbance intensity is relatively high during the switching process, so as to respond quickly to changes in system state. When the ripple fluctuation is relatively more significant, the weight of the steady-state drive duty cycle increases to avoid drastic fluctuations in the control signal due to ripple interference.

[0067] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A seamless switching control method for multiple operating modes of a UPS system, characterized in that, include: Obtain the real-time drive duty cycle sequence and DC bus voltage ripple sequence of each power switch in the UPS system under the current operating mode; The duty cycle perturbation intensity is constructed based on the frequency of sign changes of the difference between adjacent duty cycles in the real-time driving duty cycle sequence. Based on the degree of dispersion of the ratio sequence of adjacent ripple amplitudes in the DC bus voltage ripple sequence, a ripple fluctuation index is constructed. The duty cycle disturbance intensity and the ripple fluctuation index are input into the hysteresis comparator to generate a mode switching pre-trigger flag. When the mode switching pre-trigger flag is valid, extract the phase offset of the inverter voltage zero-crossing point in the previous power frequency cycle at the current moment; The output phase of the phase-locked loop is corrected based on the phase offset, and the transition drive duty cycle of each power switch is generated based on the corrected output phase. The transition drive duty cycle is weighted and averaged with the steady-state drive duty cycle of the target operating mode. The weighting coefficient is dynamically determined by the ratio of the duty cycle disturbance intensity and the ripple fluctuation index to obtain a seamless switching control signal.

2. The seamless switching control method for multi-mode operation of a UPS system according to claim 1, characterized in that, The step of constructing the duty cycle perturbation strength based on the sign change frequency of the difference between adjacent duty cycles in the real-time driving duty cycle sequence includes: The real-time drive duty cycle sequence is arranged in ascending order of sampling time, and the difference between the next duty cycle and the previous duty cycle is calculated in turn to obtain the difference sequence. The number of times the product of two adjacent differences in the difference sequence is negative is recorded as the sign inversion count. The ratio of the number of sign inversions to the total number of non-zero differences in the difference sequence is determined as the duty cycle perturbation intensity.

3. The seamless switching control method for multi-mode operation of a UPS system according to claim 1, characterized in that, The step of constructing the ripple fluctuation index based on the dispersion of the ratio sequence of adjacent ripple amplitudes in the DC bus voltage ripple sequence includes: Arrange the ripple amplitude values ​​in the DC bus voltage ripple sequence in chronological order, and calculate the ratio of the next ripple amplitude value to the previous ripple amplitude value in turn to obtain the ratio sequence. Calculate the standard deviation of all ratios in the ratio sequence to obtain the ratio standard deviation; The ratio of the standard deviation of the ratio to the mean of all ratios in the ratio sequence is determined as the ripple fluctuation index.

4. The seamless switching control method for multi-mode operation of a UPS system according to claim 1, characterized in that, The step of inputting the duty cycle disturbance intensity and the ripple fluctuation index into the hysteresis comparator to generate a mode switching pre-trigger flag includes: The difference between the duty cycle disturbance intensity and the first threshold is calculated to obtain the first deviation, and the difference between the ripple fluctuation index and the second threshold is calculated to obtain the second deviation. The first deviation and the second deviation are respectively input into two independent hysteresis comparators to obtain the first hysteresis output value and the second hysteresis output value. When both the first hysteresis output value and the second hysteresis output value are high, the mode switching pre-trigger flag is set to an active state.

5. The seamless switching control method for multi-mode operation of a UPS system according to claim 1, characterized in that, The step of extracting the phase offset of the inverter voltage zero-crossing point within the previous power frequency cycle when the mode switching pre-trigger flag is valid includes: Record the actual zero-crossing time of each inverter voltage zero-crossing point within one power frequency cycle prior to the current moment; Obtain the time difference between each actual zero-crossing moment and the corresponding ideal zero-crossing moment to obtain a zero-crossing time difference sequence; Divide each zero-crossing time difference in the zero-crossing time difference sequence by the period value of the power frequency cycle and multiply by 360 degrees to obtain the instantaneous phase shift of each zero-crossing point, and determine the average of all instantaneous phase shifts as the phase shift amount.

6. The seamless switching control method for multi-mode operation of a UPS system according to claim 1, characterized in that, The step of correcting the output phase of the phase-locked loop based on the phase offset and generating the transition drive duty cycle of each power switch based on the corrected output phase includes: The current output phase of the phase-locked loop is added to the phase offset to obtain the compensated reference phase; The switching angle sequence of each power switch is obtained by looking up the table based on the compensated reference phase; The switching angle sequence is converted into the duty cycle of each power switch in the next switching cycle to obtain the transition drive duty cycle.

7. The seamless switching control method for multi-mode operation of a UPS system according to claim 1, characterized in that, The step of weighting the transition drive duty cycle with the steady-state drive duty cycle of the target operating mode, wherein the weighting coefficient is dynamically determined by the ratio of the duty cycle disturbance intensity and the ripple fluctuation index, to obtain the seamless switching control signal, includes: Calculate the ratio of the duty cycle disturbance intensity to the ripple fluctuation index to obtain the dynamic ratio; The dynamic ratio is mapped to a closed interval between zero and one to obtain the weighting coefficients; Multiply the weighting coefficient by the transition drive duty cycle, multiply the difference between the weighting coefficient and the steady-state drive duty cycle, and add the products of the two to obtain the seamless switching control signal.

8. The seamless switching control method for multi-mode operation of a UPS system according to claim 3, characterized in that, The absolute value of each ratio in the ratio sequence is taken before the standard deviation and mean are calculated to obtain the ripple fluctuation index.

9. The seamless switching control method for multi-mode operation of a UPS system according to claim 4, characterized in that, The two independent hysteresis comparators have the same hysteresis width, and the mode switching pre-trigger flag is synchronously reset when either hysteresis output value transitions from high to low.

10. The seamless switching control method for multi-mode operation of a UPS system according to claim 7, characterized in that, The dynamic ratio is mapped to a closed interval between zero and one through a saturation function, and when the dynamic ratio exceeds a preset upper limit, the weighting coefficient is directly set to one.

Images