Modular ups inverter module fault self-test system supporting no-downtime polling detection
By constructing a cross-cycle residual vector in a modular UPS, the influence of detection residuals is identified and corrected, solving the problem of decreased comparability of detection results in non-stop polling detection of modular UPS, and improving the accuracy and stability of fault self-test results.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- GUIZHOU TAITAN ELECTRIC SYST CO LTD
- Filing Date
- 2026-05-20
- Publication Date
- 2026-06-23
AI Technical Summary
During the non-stop polling test of modular UPS, the results of the next round of testing are easily affected by the residual effects of the previous round of testing, which leads to a decrease in the comparability of the test results and misjudgment. Existing technologies are difficult to effectively identify and correct the initial observation baseline.
By collecting bus recovery parameters, power sharing status parameters, and heat dissipation status parameters, a cross-wheel residual vector is constructed to determine the detection start-up conditions or correct the initial observation baseline, delay detection, or perform baseline correction to reduce residual effects.
This improves the consistency and accuracy of fault self-test results through non-stop polling detection, reduces false positives and false negatives, and enhances the online operation and maintenance reliability of modular UPS systems.
Smart Images
Figure CN122260168A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of uninterruptible power supply (UPS) detection technology, and in particular to a modular UPS inverter module fault self-testing system that supports continuous polling detection. Background Technology
[0002] Modular UPS systems typically consist of multiple inverter modules operating in parallel, all connected to a common bus and supplying power to the load. To ensure continuous power supply, it is generally desirable to perform fault self-checks on each inverter module sequentially without shutting down the system.
[0003] In existing non-stop polling testing schemes, the testing process for the next inverter module typically starts directly after the current inverter module under test has finished testing, based on conditions such as the recovery of the common bus voltage, the recovery of the total output current, or the end of a preset waiting period. However, in a modular UPS operating in parallel, multiple inverter modules typically undergo dynamic processes such as current sharing adjustment, power redistribution, integral accumulation of control loops, heat dissipation adjustment, and temperature rise hysteresis. Even if the surface parameters of the common bus have recovered to the allowable range, residual states may still exist within the system, causing a shift in the starting observation baseline for the next inverter module.
[0004] In this situation, the detection results of the next inverter module include not only the inverter module's own response information but also baseline offset information caused by the residual effects of the previous round of detection. This can easily lead to a decrease in the comparability of detection results from different rounds, and may even result in misjudging the residual effects of the previous round as an anomaly of the currently tested inverter module. Therefore, how to identify the residual effects of the previous round of detection on the starting state of the next round of detection during the non-stop polling detection of modular UPS, and on this basis determine the timing of starting the next round of detection or correct the starting observation baseline of the next round, has become a technical problem that needs to be solved in this field. Summary of the Invention
[0005] This application provides a modular UPS inverter module fault self-testing system that supports non-stop polling detection, in order to improve the comparability and reliability of fault self-testing results of adjacent cycles of inverter modules.
[0006] In the first aspect, this application provides a modular UPS inverter module fault self-testing system that supports non-stop polling detection, including a management device, multiple inverter modules connected in parallel, a data acquisition unit, and a heat dissipation unit. The data acquisition unit is used to acquire the bus recovery parameters, power sharing status parameters and heat dissipation status parameters of the modular UPS after the current inverter module under test has finished testing. The management device is used to construct a cross-cycle residual vector characterizing the residual impact of the previous round of detection based on the bus recovery parameters, the power sharing status parameters, and the heat dissipation status parameters. It is also used to determine, based on the residual vector across the wheel, whether the detection start-up conditions of the next tested inverter module have been met; It is also used to delay the start of the detection of the next inverter module under test when the detection start condition is not met, or to correct the starting observation baseline of the next inverter module under test according to the cross-wheel residual vector; It is also used to control the next inverter module under test to perform a fault self-test after the detection start condition is met or the initial observation baseline correction is completed, and to output the self-test conclusion of the next inverter module under test.
[0007] In the above system, the management device does not determine whether to proceed to the next round of testing solely based on the surface recovery status of the common bus. Instead, it further combines the power sharing status and heat dissipation status to construct a cross-bus residual vector, characterizing the residual effects that have not yet subsided inside the modular UPS after the previous round of testing. Through this approach, the management device can delay starting the next round of testing before the residual effects from the previous round have completely subsided, or correct the starting observation baseline of the next tested inverter module. This reduces the impact of cross-bus contamination on the results of the next round of testing, improving the consistency and accuracy of inverter module fault self-test results in non-stop polling testing scenarios.
[0008] In one possible design, the method further includes: after the next inverter module under test completes its fault self-test, treating the next inverter module under test as the new current inverter module under test, continuing to collect bus recovery parameters, power sharing status parameters, and heat dissipation status parameters after the test, to construct a cross-wheel residual vector corresponding to the subsequent inverter module under test, and performing a test start judgment for the subsequent inverter module under test. Through this design, the management device can repeatedly perform cross-wheel residual identification and start condition judgment during the polling test of multiple inverter modules, thereby achieving continuous, non-stop polling self-testing of multiple inverter modules.
[0009] In one possible design, when the next inverter module under test includes multiple candidate inverter modules, the management device first determines the next inverter module under test based on priority, and then judges the determined next inverter module under test, that is, it judges based on the cross-round residual vector to execute the detection start condition. Thus, the management device can make orderly selections among multiple candidate inverter modules under test to optimize the polling detection strategy.
[0010] In one possible design, the priority of the plurality of candidate inverter modules under test is related to the health risk information of the plurality of candidate inverter modules under test; wherein, the higher the risk level indicated by the health risk information, the higher the priority of the candidate inverter module under test corresponding to the health risk information. Optionally, the health risk information can be determined based on at least one of the following: historical failure count, number of warnings, fault self-inspection conclusions of the most recent preset round, cumulative temperature rise anomaly, and cumulative current sharing deviation.
[0011] In one possible design, the priority of the plurality of candidate inverter modules under test is related to the power sharing information of the plurality of candidate inverter modules under test; wherein, the smaller the power sharing indicated by the power sharing information, the higher the priority of the candidate inverter module under test corresponding to the power sharing information. Optionally, the power sharing information may be the average power sharing within the current sampling window, or the average power sharing within the most recent preset time period.
[0012] In one possible design, the priority of the plurality of candidate inverter modules under test is related to their thermal margin information; wherein, the larger the thermal margin indicated by the thermal margin information, the higher the priority of the candidate inverter module under test corresponding to the thermal margin information. Optionally, the thermal margin information can be determined based on the difference between the upper temperature limit and the current module temperature and / or the remaining fan speed adjustment headroom.
[0013] In one possible design, the cross-wheel residual vector includes a bus residual sub-vector, a power sharing residual sub-vector, and a heat dissipation residual sub-vector; wherein, the bus residual sub-vector is determined according to the bus recovery parameters, the power sharing residual sub-vector is determined according to the power sharing state parameters, and the heat dissipation residual vector is determined according to the heat dissipation state parameters.
[0014] In one possible design, determining whether the detection start-up condition for the next inverter module under test is met based on the residual vector of the cross-wheel includes: determining whether the residual sub-vector of the bus, the residual sub-vector of the power sharing, and the residual sub-vector of the heat dissipation meet the corresponding residual threshold conditions; and determining that the detection start-up condition is met when the residual sub-vector of the bus, the residual sub-vector of the power sharing, and the residual sub-vector of the heat dissipation all meet the corresponding residual threshold conditions.
[0015] In one possible design, correcting the initial observation baseline of the next inverter module under test based on the residual vector across the wheel includes: Based on the residual vector across the wheel, determine the residual correction value of at least one observation parameter corresponding to the next inverter module under test; based on the initial observation value of the at least one observation parameter and the residual correction value, determine the corrected starting observation baseline of the next inverter module under test.
[0016] In one possible design, when the detection start condition is not met, the step of correcting the starting observation baseline of the next inverter module under test based on the cross-wheel residual vector includes: after the delayed start time reaches a preset time, determining the residual correction value of the at least one observation parameter based on the cross-wheel residual vector at the current moment and the pre-stored residual attenuation relationship; and starting the fault self-test of the next inverter module under test based on the corrected starting observation baseline.
[0017] The technical effects and advantages of this invention's modular UPS inverter module fault self-testing system that supports non-stop polling detection are as follows: This invention constructs a cross-round residual vector characterizing the residual impact of the previous round of testing by comprehensively collecting bus recovery parameters, power sharing status parameters, and heat dissipation status parameters after the current inverter module under test has finished testing. Based on this vector, the timing of starting the next round of testing for the next inverter module under test can be determined, or the initial observation baseline of the next inverter module under test can be corrected. This avoids the problem of prematurely starting the next round of testing based solely on the surface recovery status of the common bus, reduces the contamination effect of the previous round of testing residue on the results of the next round of fault self-testing, and improves the comparability and consistency between the results of different rounds of testing. At the same time, it enables continuous polling testing of multiple parallel inverter modules under the condition of uninterrupted power supply, which helps to reduce misjudgments and omissions, improves the accuracy and stability of the inverter module fault self-testing conclusions, and enhances the reliability of online operation and maintenance of modular UPS systems. Attached Figure Description
[0018] Figure 1 This is a schematic diagram of the architecture of a modular UPS system provided in an embodiment of this application.
[0019] Figure 2 This is a schematic diagram illustrating a process for constructing a wheel-crossing residual vector, as provided in an embodiment of this application.
[0020] Figure 3 The flowchart illustrates a modular UPS inverter module fault self-testing method that supports non-stop polling detection, as provided in an embodiment of this application.
[0021] Figure 4 This is a schematic diagram of the hardware structure of a management device provided in an embodiment of this application. Detailed Implementation
[0022] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the application will be further described in detail below with reference to the embodiments. The terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.
[0023] Figure 1This is a schematic diagram of the architecture of a modular UPS system provided in an embodiment of this application. Figure 4 As shown, the system may include a management device, multiple inverter modules connected in parallel, a common bus, and a load. Optionally, the system may also include a heat dissipation unit and a data acquisition unit. The management device is used to acquire the operating parameters of each inverter module, schedule the polling detection process of each inverter module, and output corresponding fault self-check conclusions.
[0024] In this embodiment, multiple inverter modules are connected in parallel to a common bus and jointly supply power to the load. Each inverter module may have output voltage regulation, current sharing control, and fault self-diagnosis functions. During parallel power supply, the multiple inverter modules can dynamically share power according to the current load conditions. The heat dissipation unit may include at least one of a fan, a radiator, or a liquid-cooled component. The data acquisition unit can be used to collect bus recovery parameters, power sharing status parameters, and heat dissipation status parameters.
[0025] Among them, the bus recovery parameters may include at least one of the following: bus voltage deviation, bus current fluctuation, and bus ripple; the power sharing status parameters may include at least one of the following: the sharing current parameter, sharing power parameter, and current sharing deviation parameter of each inverter module; and the heat dissipation status parameters may include at least one of the following: inverter module temperature parameter, radiator temperature parameter, and fan operating parameter.
[0026] In a non-stop polling testing scenario, after the testing of the current inverter module under test is completed, the surface parameters of the common bus may have returned to the allowable range. However, the power sharing relationship, control loop status, and heat dissipation status among multiple inverter modules may still be in a transition process. If the testing of the next inverter module under test is started directly before the transition process is over, the starting observation baseline of the next inverter module under test may have been affected by the residual influence of the previous round of testing, thereby reducing the comparability between the test results of different rounds.
[0027] Based on the above problems, this application provides a modular UPS inverter module fault self-testing method that supports non-stop polling detection. This method can be... Figure 1 The management device shown can be executed in conjunction with the controller in the modular UPS. This application does not limit the specific physical form of the management device; it can be a UPS control board, a standalone monitoring host, or a server device.
[0028] It should be noted that the specific application scenarios of the modular UPS are not limited in the embodiments of this application. The method can be applied to data center power systems, industrial uninterruptible power supply systems, rail transit power supply systems and other scenarios that use multiple inverter modules connected in parallel for power supply.
[0029] The following describes the modular UPS inverter module fault self-testing method that supports non-stop polling detection provided in the embodiments of this application.
[0030] Figure 3 A flowchart illustrating a modular UPS inverter module fault self-testing method supporting non-stop polling detection, provided as an embodiment of this application. (Reference) Figure 3 The method may include the following steps.
[0031] S301: After the current inverter module under test is completed, the management device collects the bus recovery parameters, power sharing status parameters and heat dissipation status parameters of the modular UPS.
[0032] In the embodiments of this application, such as Figure 2 As shown, the inverter module currently under test can be the one that has just completed the previous round of fault self-testing. After the current inverter module under test finishes testing, the management device does not directly determine whether to proceed to the next round of testing based on a single waiting time or whether the bus voltage has returned to the set range. Instead, it continues to collect multiple parameters related to cross-cycle residual loads. Among these, the bus recovery parameter reflects whether the common bus has recovered from the disturbance caused by the previous round of testing; the power sharing status parameter reflects whether the power redistribution among multiple inverter modules is still in progress; and the heat dissipation status parameter reflects whether the heat accumulation caused by the previous round of testing has subsided.
[0033] For example, within a preset sampling window after the current inverter module under test has finished testing, the management device can collect data at preset time intervals, including the common bus voltage deviation, the current shared by each inverter module, the power shared by each inverter module, the inverter module temperature, and the fan speed. This preset sampling window can be set to a time range of hundreds of milliseconds to several seconds according to the system configuration; this embodiment does not limit the time range.
[0034] In this embodiment of the application, while collecting the bus recovery parameters, power sharing status parameters and heat dissipation status parameters, the management device can also detect whether there is a sudden change in external operating conditions within the current sampling window; The external operating condition change event may include at least one of the following: load change event, rapid change in ambient temperature event, and fan control mode switching event.
[0035] In one embodiment, the management device can determine whether an external operating condition abrupt event has occurred based on at least one of the following: bus current change rate, load power change rate, ambient temperature change rate, and fan control status change flag. When the external operating condition abrupt event occurs, the management device can mark the current sampling window as an invalid residual judgment window.
[0036] S302: The management device constructs a cross-cycle residual vector characterizing the residual impact of the previous round of detection based on the bus recovery parameters, the power sharing status parameters, and the heat dissipation status parameters.
[0037] In this embodiment of the application, the cross-wheel residual vector may include a bus residual sub-vector, a power sharing residual sub-vector, and a heat dissipation residual sub-vector; The busbar residual sub-vector can be determined based on the busbar recovery parameters and is used to characterize the residual impact on the common busbar side; The residual subvector of power sharing can be determined based on the power sharing state parameters and is used to characterize the degree to which the power redistribution among multiple inverter modules has not yet converged. The residual heat dissipation subvector can be determined based on the heat dissipation state parameters and is used to characterize the residual effects caused by heat hysteresis and heat dissipation inertia.
[0038] In this embodiment of the application, when the management device constructs the wheel-crossing residual vector, it can first classify and standardize the collected bus recovery parameters, power sharing status parameters and heat dissipation status parameters, and then generate bus residual sub-vectors, power sharing residual sub-vectors and heat dissipation residual sub-vectors respectively. The wheel-crossing residual vector is formed by combining the bus residual sub-vectors, the power sharing residual sub-vectors and the heat dissipation residual sub-vectors.
[0039] In one embodiment, the bus residual subvector can be composed of bus voltage deviation and bus ripple increment; the power sharing residual subvector can be composed of power sharing deviation and current sharing deviation of each inverter module; and the heat dissipation residual subvector can be composed of module temperature deviation and fan speed deviation corresponding to the next inverter module under test.
[0040] In this default implementation, the management device can divide the bus voltage deviation by a preset bus allowable fluctuation threshold and the bus ripple increment by a preset ripple allowable threshold to obtain each normalized component in the bus residual sub-vector; divide the power sharing deviation by a preset power sharing deviation threshold and the current sharing deviation by a preset current sharing deviation threshold to obtain each normalized component in the power sharing residual sub-vector; and divide the module temperature deviation by a preset temperature rise deviation threshold and the fan speed deviation by a preset speed deviation threshold to obtain each normalized component in the heat dissipation residual sub-vector.
[0041] In this default implementation, the management device can first calculate the sub-scores corresponding to the bus residual sub-vector, power sharing residual sub-vector, and heat dissipation residual sub-vector respectively, and then perform a weighted summation of the sub-scores according to preset weights to obtain a comprehensive residual score. The comprehensive residual score and the three sub-vectors are used together as the representation result of the cross-wheel residual vector. For example, the preset weights can be set to 0.3, 0.4, and 0.3 in the order of bus residual, power sharing residual, and heat dissipation residual.
[0042] S303: The management device determines whether the detection start condition for the next inverter module under test has been met based on the residual vector of the cross wheel.
[0043] In this embodiment of the application, the detection start conditions may include direct start conditions and modified start conditions; The direct initiation conditions are used to determine whether the current residual impact has been reduced to a level that allows direct entry into the next round of testing; The modified initiation conditions are used to determine that although the current residual effects have not completely subsided, they are still within the range that can be compensated for by baseline correction.
[0044] In one embodiment, when the bus residual sub-vector, the power sharing residual sub-vector, and the heat dissipation residual sub-vector all satisfy their respective first threshold conditions, the management device can determine that the direct start condition has been met; when at least one sub-vector does not satisfy the corresponding first threshold condition, but the sub-vector that does not satisfy the first threshold condition still satisfies the corresponding second threshold condition, the management device can determine that the corrected start condition has been met; wherein, the allowable residual range corresponding to the second threshold condition is greater than the allowable residual range corresponding to the first threshold condition.
[0045] In this embodiment, the first threshold condition, the second threshold condition, and the fault determination threshold can be uniformly maintained by the same threshold management unit. The threshold management unit can store parameter threshold configuration tables corresponding to each inverter module, and these tables include at least bus recovery thresholds, power sharing thresholds, residual heat dissipation thresholds, and fault determination thresholds.
[0046] In one implementation, the first threshold condition, the second threshold condition, and the fault determination threshold can be determined based on at least one of the rated parameters of the modular UPS, factory calibration results, and historical stable operation data. For example, the first threshold condition can be less than the second threshold condition, and the second threshold condition can be less than or equal to the safety upper bound of the corresponding fault determination threshold, thereby maintaining a consistent threshold hierarchy among direct start determination, corrective start determination, and fault determination.
[0047] Optionally, the management device can also adjust the first threshold condition and the second threshold condition based on the current load rate, ambient temperature, and operating mode.
[0048] When the current sampling window is marked as an invalid residual judgment window, the management device can extend the residual observation time and re-collect the bus recovery parameters, power sharing status parameters and heat dissipation status parameters after the external operating conditions stabilize again, so as to re-execute the direct start judgment and the corrected start judgment.
[0049] If multiple candidate inverter modules exist, the management device can first determine the priority of the multiple candidate inverter modules, and then determine the next inverter module to be tested in descending order of priority. The priority of multiple candidate inverter modules can be related to at least one of the following: health risk information, power sharing information, and thermal margin information.
[0050] The health risk information can be determined based on at least one of the following: historical fault count, warning count, fault self-inspection conclusion of the most recent preset round, cumulative temperature rise abnormality, and cumulative current sharing deviation; the power sharing information can be the average power sharing within the current sampling window or the average power sharing within the most recent preset time period; the heat dissipation margin information can be determined based on the difference between the upper temperature limit and the current module temperature and / or the remaining fan speed adjustment space.
[0051] In a default implementation, the management device can preferentially select candidate inverter modules under test that meet the direct start conditions; when multiple candidate inverter modules under test all meet the direct start conditions, the management device then determines the next inverter module under test in descending order of priority of the multiple candidate inverter modules under test; when multiple candidate inverter modules under test do not meet the direct start conditions but at least one candidate inverter module under test meets the modified start conditions, the management device can determine the next inverter module under test from the candidate inverter modules under test that meet the modified start conditions in descending order of priority.
[0052] S304: If the detection start condition is not met, the management device delays the start of the detection of the next inverter module under test, or corrects the starting observation baseline of the next inverter module under test according to the cross-wheel residual vector.
[0053] In this embodiment of the application, the management device can collect bus recovery parameters, power sharing status parameters and heat dissipation status parameters again after a preset delay when the detection start conditions are not met, and reconstruct the cross-wheel residual vector at the current moment.
[0054] In a default implementation, the residual attenuation relationship can be obtained by the management device during the factory commissioning or maintenance calibration phase of the modular UPS. Specifically, the management device can sequentially trigger the sampling process after the current inverter module under test finishes testing under multiple preset load rate conditions, and record the changes of bus recovery parameters, power sharing status parameters, and heat dissipation status parameters over time, thereby generating residual attenuation sample data corresponding to each parameter.
[0055] The management device can establish residual attenuation relationships for at least one of the output voltage parameter, output current parameter, module temperature parameter, and ripple parameter based on the residual attenuation sample data. For example, the residual attenuation relationship can be in the form of a mapping table stored in segments by time period, where the mapping table records the residual retention ratio or residual correction ratio under different delay durations.
[0056] Optionally, during system operation, the management device can also update the residual attenuation relationship based on historical polling detection data; however, before the update is completed, the management device will prioritize using the residual attenuation relationship pre-stored during the factory commissioning stage or the maintenance calibration stage.
[0057] In this embodiment, when the management device corrects the starting observation baseline of the next inverter module under test based on the residual vector across the wheel, it can determine the corresponding residual correction value for at least one observation parameter corresponding to the next inverter module under test. The at least one observation parameter may include at least one of the following: output voltage observation value, output current observation value, ripple observation value, and module temperature observation value.
[0058] In one implementation, the management device determines the residual correction value of the target observation parameter based on the residual component in the cross-wheel residual vector corresponding to the target observation parameter, combined with the residual attenuation relationship corresponding to the target observation parameter. Optionally, the residual correction value can be obtained by a lookup table method, a mapping relationship method, or a proportional calculation method.
[0059] In one implementation, the management device can determine the corrected initial observation baseline based on the following relationship: The corrected initial observation baseline is equal to the currently acquired initial observation value minus the corresponding positive residual correction value; Alternatively, the corrected initial observation baseline is equal to the currently acquired initial observation value plus the corresponding negative residual correction value.
[0060] S305: After the detection start condition is met or the initial observation baseline correction is completed, the management device performs a fault self-test on the next inverter module under test and outputs the self-test conclusion of the next inverter module under test.
[0061] In this embodiment, when the management device performs a fault self-test on the next inverter module under test, it can apply detection process control to the next inverter module under test based on the corrected initial observation baseline, and collect the response data of the next inverter module under test within the detection window. The detection process control may include keeping the next inverter module under test running under a preset output condition, or applying a small detection disturbance of a preset amplitude to the next inverter module under test.
[0062] In one implementation, the management device can apply a limited-amplitude detection disturbance to the next inverter module under test without changing the continuous power supply state of the modular UPS. The detection disturbance can be at least one of a small change in the output reference quantity, a small change in the duty cycle, or a small change in the control command.
[0063] In another implementation, the management device may not apply additional detection disturbances to the next inverter module under test. Instead, after the next inverter module under test enters the preset detection window, it directly collects the natural response data of the next inverter module under test under the current power supply conditions and compares the natural response data with the corrected initial observation baseline to determine whether the next inverter module under test has any abnormalities.
[0064] In this embodiment, the management device can determine at least one fault determination index based on the response data and the corrected initial observation baseline. The fault determination index may include at least one of the following: output voltage deviation, output current deviation, ripple increment, temperature rise change, and current sharing deviation.
[0065] In one implementation, when at least one fault determination indicator exceeds a corresponding determination threshold, the management device can determine that the next inverter module under test is faulty; when the number of indicators exceeding the corresponding determination threshold among multiple fault determination indicators reaches a preset number, the management device can determine that the next inverter module under test is faulty. Optionally, when outputting the fault self-test conclusion, the management device can call the parameter threshold configuration table shared with the start condition judgment to ensure that the initial baseline correction process and the fault determination process adopt a consistent parameter boundary system.
[0066] For example, the management device can compare the output voltage deviation with a first fault threshold, the ripple increment with a second fault threshold, and the current sharing deviation with a third fault threshold. When at least two of the output voltage deviation, ripple increment, and current sharing deviation exceed their corresponding fault thresholds, the management device can output a fault self-test conclusion of "abnormal" for the next inverter module under test; when only one exceeds its corresponding fault threshold, the management device can output a fault self-test conclusion of "warning" for the next inverter module under test.
[0067] After the next inverter module under test completes its fault self-test, the management device can use the next inverter module under test as the new current inverter module under test and continue to execute steps S301 to S305 to perform continuous non-stop polling tests on subsequent inverter modules.
[0068] Optionally, after each round of testing, the management device can also store the cross-round residual vector, residual correction value and final self-test conclusion corresponding to that round of testing, so as to update the pre-stored residual attenuation relationship or optimize the testing strategy for different inverter modules.
[0069] The following example illustrates the cross-vessel residual identification and baseline correction process provided in this application. Assume a modular UPS includes four parallel inverter modules M1, M2, M3, and M4, which together supply power to the load. The rated bus voltage is 400 volts. The inverter module currently under test is M1, and the next candidate inverter modules to be tested include M2, M3, and M4.
[0070] After M1 completes its current round of fault self-check, the management device collects relevant parameters within a one-second sampling window. The bus voltage deviation collected by the management device is three volts, and the bus ripple increment is 0.8 volts; the power sharing deviations of M2, M3, and M4 are three percent, two percent, and one percent, respectively; the module temperature deviation of M2 is two degrees Celsius, and the fan speed deviation is 120 revolutions per minute.
[0071] The management equipment normalizes the bus voltage deviation and bus ripple increment to form a bus residual sub-vector; normalizes the power sharing deviation of each inverter module to form a power sharing residual sub-vector; and normalizes the module temperature deviation and fan speed deviation to form a heat dissipation residual sub-vector.
[0072] Suppose the management device's preset direct start conditions are: the score corresponding to the bus residual sub-vector is no greater than 0.3, the score corresponding to the power sharing residual sub-vector is no greater than 0.4, and the score corresponding to the heat dissipation residual sub-vector is no greater than 0.3; the preset corrected start conditions are: the score corresponding to the bus residual sub-vector is no greater than 0.5, the score corresponding to the power sharing residual sub-vector is no greater than 0.6, and the score corresponding to the heat dissipation residual sub-vector is no greater than 0.5. Based on the current sampled data, the management device calculates that the score corresponding to the bus residual sub-vector is 0.28, the score corresponding to the power sharing residual sub-vector is 0.35, and the score corresponding to the heat dissipation residual sub-vector is 0.42. At this point, the management device can determine that the current state does not meet the direct start conditions, but meets the corrected start conditions.
[0073] Assuming M2 has a higher health risk level than M3 and M4, and its current power contribution is lower than that of M3, the management device can identify M2 as the next inverter module to be tested. The management device then determines the residual correction value for M2's temperature parameter to be 1.5 degrees Celsius and the residual correction value for its output current parameter to be 0.2 amperes, based on the current residual heat dissipation subvector and the pre-stored residual attenuation relationship. If the initial observed temperature of M2 before the start of testing is 45 degrees Celsius and the initial observed output current is 12 amperes, the management device can determine 43.5 degrees Celsius and 11.8 amperes as the corrected starting temperature baseline and corrected starting current baseline for M2, respectively.
[0074] Subsequently, the management device performs a fault self-check on M2. Assuming that within the detection window, the temperature rise of M2 relative to the corrected starting temperature baseline is 0.6 degrees Celsius, the current deviation relative to the corrected starting current baseline is 0.1 amperes, and the output voltage deviation relative to the corrected starting voltage baseline does not exceed the corresponding fault threshold, then the management device determines that M2 is not abnormal. If at least two of the temperature rise change and current deviation exceed the corresponding fault threshold, then the management device outputs a fault self-check conclusion for M2 as abnormal.
[0075] After M2 completes its fault self-check, the management device can use M2 as the new currently tested inverter module to continue performing cross-cycle residual identification and startup judgment for subsequent inverter modules. As can be seen from the above example, the method provided in this application embodiment can identify the impact of the previous round of detection residuals on the starting state of the next round of detection during non-stop polling detection, and correct the baseline of the next round of detection when necessary, thereby improving the comparability between fault self-check results from different rounds.
Claims
1. A modular UPS inverter module fault self-testing system supporting non-stop polling detection, characterized in that, It includes management equipment, multiple inverter modules connected in parallel, a data acquisition unit, and a heat dissipation unit; The data acquisition unit is used to acquire the bus recovery parameters, power sharing status parameters and heat dissipation status parameters of the modular UPS after the current inverter module under test has finished testing. The management device is used to construct a cross-cycle residual vector characterizing the residual impact of the previous round of detection based on the bus recovery parameters, the power sharing status parameters, and the heat dissipation status parameters. It is also used to determine, based on the residual vector across the wheel, whether the detection start-up conditions of the next tested inverter module have been met; It is also used to delay the start of the detection of the next inverter module under test when the detection start condition is not met, or to correct the starting observation baseline of the next inverter module under test according to the cross-wheel residual vector; It is also used to control the next inverter module under test to perform a fault self-test after the detection start condition is met or the initial observation baseline correction is completed, and to output the self-test conclusion of the next inverter module under test.
2. The modular UPS inverter module fault self-testing system supporting non-stop polling detection according to claim 1, characterized in that: The management device is also used to, after the next inverter module under test completes its fault self-test, treat the next inverter module under test as the new current inverter module under test, and continue to control the data acquisition unit to collect the bus recovery parameters, power sharing status parameters and heat dissipation status parameters after the test is completed, so as to construct the cross-wheel residual vector corresponding to the subsequent inverter module under test, and execute the test start judgment of the subsequent inverter module under test.
3. The modular UPS inverter module fault self-testing system supporting non-stop polling detection according to claim 1, characterized in that: When the next inverter module under test includes multiple candidate inverter modules under test, the management device is used to determine the next inverter module under test in descending order of priority of the multiple candidate inverter modules under test, and to determine whether the determined next inverter module under test meets the detection start condition.
4. The modular UPS inverter module fault self-testing system supporting non-stop polling detection according to claim 3, characterized in that: The priority of the plurality of candidate inverter modules under test is related to the health risk information of the plurality of candidate inverter modules under test; The higher the risk level indicated by the health risk information, the higher the priority of the candidate inverter module to be tested corresponding to the health risk information.
5. The modular UPS inverter module fault self-testing system supporting non-stop polling detection according to claim 3 or 4, characterized in that, The priority of the plurality of candidate inverter modules under test is related to the power sharing information of the plurality of candidate inverter modules under test; The smaller the power sharing indicated by the power sharing information, the higher the priority of the candidate inverter module under test corresponding to the power sharing information.
6. The modular UPS inverter module fault self-testing system supporting non-stop polling detection according to claim 3 or 4, characterized in that: The priority of the multiple candidate inverter modules under test is related to the heat dissipation margin information of the multiple candidate inverter modules under test. The larger the heat dissipation margin indicated by the heat dissipation margin information, the higher the priority of the candidate inverter module under test corresponding to the heat dissipation margin information.
7. The modular UPS inverter module fault self-testing system supporting non-stop polling detection according to claim 2, characterized in that: The residual vector across the wheel includes a bus residual sub-vector, a power sharing residual sub-vector, and a heat dissipation residual vector; The bus residual sub-vector is determined based on the bus recovery parameters, the power sharing residual sub-vector is determined based on the power sharing state parameters, and the heat dissipation residual sub-vector is determined based on the heat dissipation state parameters.
8. The modular UPS inverter module fault self-testing system supporting non-stop polling detection according to claim 7, characterized in that: The management device is used to determine whether the bus residual sub-vector, the power sharing residual sub-vector, and the heat dissipation residual sub-vector meet the corresponding residual threshold conditions; And when the bus residual sub-vector, the power sharing residual sub-vector, and the heat dissipation residual sub-vector all satisfy the corresponding residual threshold conditions, it is determined that the detection start condition has been met.
9. The modular UPS inverter module fault self-testing system supporting non-stop polling detection according to claim 8, characterized in that: The management device is also used to determine the residual correction value of at least one observation parameter corresponding to the next inverter module under test based on the residual vector of the cross-wheel; Based on the initial observation value of the at least one observation parameter and the residual correction value, the corrected starting observation baseline of the next inverter module under test is determined.
10. The modular UPS inverter module fault self-testing system supporting non-stop polling detection according to claim 9, characterized in that: When the detection start condition is not met, the management device is also used to determine the residual correction value of the at least one observation parameter based on the current wheel residual vector and the pre-stored residual decay relationship after the delayed start time reaches the preset time. Based on the corrected initial observation baseline, the next inverter module under test is controlled to perform a fault self-test.