A modular power management system with embedded health assessment

Abstract

The application discloses a modular power management system for embedded health state evaluation and belongs to the technical field of power electronic equipment detection, and comprises a main controller and a plurality of power modules, wherein the power module comprises a cooperative processing unit and a boundary control unit controlled by the cooperative processing unit. By injecting a perturbation instruction containing time limit and amplitude limit to a running module and synchronously analyzing bidirectional dynamic response data, dynamic stability margin and cooperative compensation capacity of an online quantification system can be realized, and a high-confidence health trend spectrum is further constructed.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power electronic device detection, in particular to a modular power management system with embedded health state evaluation. BACKGROUND

[0002] Modular power supply systems are widely used in industrial control, data center construction, communication base station power supply and other fields due to their flexible expansion and improved reliability through redundant configuration. In such systems, ensuring that each power conversion module maintains good health is crucial to ensuring the continuous operation of the entire mission-critical facility. Therefore, equipping the system with a corresponding monitoring system to monitor the internal state and remaining life of the power supply equipment in real time has become a core requirement for related product design and maintenance.

[0003] Existing power supply system monitoring solutions usually use static parameter-based monitoring methods, that is, by continuously reading the DC voltage, DC current and temperature values of the internal heat sink of the module through sensors. When the monitoring system finds that the collected values exceed the pre-set upper and lower threshold alarm lines, it determines that there is an abnormality or failure in the channel and then cuts off the channel. Some more advanced methods count the cumulative running time of the power module and the historical maximum temperature it has experienced, and use an aging mathematical model of electronic components to estimate the theoretical degree of life decay. In addition, there are methods that periodically disconnect the power module from the system bus and connect it to an external test load box for a complete charge and discharge test or step load test.

[0004] However, traditional static monitoring methods cannot reflect the early drift of internal parameters in the power control loop. When the module control capability decreases but the static output remains within the stable band, the monitoring system often misjudges its state as good, lacking the ability to identify the implicit degradation of insufficient dynamic stability reserves. The life estimated by relying on mathematical models does not take into account the manufacturing differences and actual dynamic working conditions of specific individuals, resulting in a large error. The off-line load test method is complex and interrupts the online state of the module, which is not suitable for applications with high continuity requirements. More importantly, existing methods lack an evaluation mechanism for the rapid collaborative response capability of multiple modules when running in parallel, and cannot reveal potential weak links when one individual fluctuates whether the remaining modules have qualified compensation and support capabilities. SUMMARY

[0005] To solve the above problems, the present application provides a modular power management system with embedded health state evaluation method, which can realize online quantification of the dynamic stability margin and collaborative compensation capability of the system by injecting perturbation instructions containing time and amplitude restrictions into the running module and synchronously analyzing the bidirectional dynamic response data, and further constructing a high-confidence health trend map.

[0006] The above object can be achieved by the following solution:

[0007] A modular power management system for embedded health state evaluation, comprising a main controller, a plurality of power modules, the power module comprising a cooperative processing unit and a boundary control unit controlled by the cooperative processing unit; wherein the main controller is configured to determine whether the current system operating state meets the preset perturbation test safety condition; wherein the perturbation test safety condition comprises that the system is in a steady state working condition, the total load rate of the system is lower than a safety threshold, and there is no fault alarm within a preset time length; if yes, a predetermined perturbation test sequence is followed to select one of the plurality of power modules as a current tested module; a boundary perturbation instruction is sent to the cooperative processing unit of the current tested module, and the cooperative processing unit of the current tested module controls the boundary control unit of the current tested module to perform a performance boundary adjustment operation within a preset perturbation duration; wherein the boundary perturbation instruction comprises a voltage perturbation instruction or a current limit perturbation instruction; within the perturbation duration, first type response data is obtained from the cooperative processing unit of the current tested module, and a dynamic stability margin index of the current tested module is analyzed and calculated based on the first type response data; within the perturbation duration, second type response data is obtained from the cooperative processing unit of at least one other power module, and a cooperative compensation capability index of the system is analyzed and calculated based on the second type response data; the dynamic stability margin index and the cooperative compensation capability index are fused to generate a health potential index of the current tested module.

[0008] Based on the same inventive concept, the application also provides a modular power management method for embedded health state evaluation, which comprises: a main controller determining whether a current system running state meets a preset perturbation test safety condition; wherein the perturbation test safety condition comprises that the system is in a steady state working condition, the total load rate of the system is lower than a safety threshold, and there is no fault alarm within a preset time length; if yes, the main controller selects one from a plurality of power modules as a current tested module according to a predetermined perturbation test sequence; the main controller sends a boundary perturbation instruction to a cooperative processing unit of the current tested module, and the cooperative processing unit of the current tested module controls a boundary control unit of the current tested module to perform a performance boundary adjustment operation within a preset perturbation duration; wherein the boundary perturbation instruction comprises a voltage perturbation instruction or a current limit perturbation instruction; within the perturbation duration, the main controller acquires first type response data from the cooperative processing unit of the current tested module, and analyzes and calculates a dynamic stability margin index of the current tested module based on the first type response data; within the perturbation duration, the main controller acquires second type response data from the cooperative processing unit of at least one other power module, and analyzes and calculates a cooperative compensation capability index of the system based on the second type response data; and the main controller fuses the dynamic stability margin index and the cooperative compensation capability index to generate a health potential index of the current tested module.

[0009] Compared with the prior art, the application has the following advantages:

[0010] By applying imperceptible voltage or current perturbation instructions when the power module is normally running and in a safe state, the application can actively stimulate and capture the dynamic response characteristics of the power module and the system without interrupting power supply and introducing a large power test load. This embedded perturbation test mechanism enables technicians to directly obtain the feedback loop gain and damping state of the power supply circuit, breaking through the hysteresis limitation of the traditional passive monitoring method which can only alarm faults according to voltage or current deviation from the rated value, improving the real-time performance and sensitivity of the power system health state evaluation, and realizing the transition from fault alarm to early degradation warning.

[0011] The embodiment of the application not only analyzes the dynamic stability margin of the tested module itself, but also proposes and calculates the cooperative compensation capability index of the system. By monitoring the current support or voltage suppression capability of the non-tested module when it receives fluctuations on the bus, the scheme quantifies and evaluates the robustness of the parallel system as a whole. This evaluation system considers the interaction between modules and can find the system-level oscillation risk caused by parameter mismatching or aging in parallel networks although the single machine test is qualified, thereby ensuring the power supply quality and stability of the parallel operation of multiple modules.

[0012] The application introduces a health trend correction coefficient by combining the potential index change rate obtained by short-term perturbation test and the state evaluation change rate based on long-term operation data. The health state evaluation value of the power module is adaptively corrected by combining the physical control characteristic change of the fast time scale and the environmental stress accumulation effect of the slow time scale. This effectively corrects the deviation of the theoretical calculation based on the running time, so that the maintenance system can optimize the load distribution and standby module enabling strategy according to the more accurate corrected evaluation value, maximize the actual service life of the power equipment and reduce the maintenance cost. BRIEF DESCRIPTION OF DRAWINGS

[0013] Figure 1 is a structural schematic diagram of an embedded health state evaluation modular power management system according to an embodiment of the application.

[0014] Figure 2 is a flowchart of an embedded health state evaluation modular power management method according to an embodiment of the application. DETAILED DESCRIPTION

[0015] To make the objectives, technical solutions and advantages of the embodiments of the application clearer, the technical solutions in the embodiments of the application will be described below in detail with reference to the drawings of the embodiments of the application. Obviously, the described embodiments are some but not all of the embodiments of the application. Based on the embodiments in the application, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the protection scope of the application.

[0016] With reference to Figure 1 An embodiment of the application provides an embedded health state evaluation modular power management system. By injecting a perturbation instruction containing aging and amplitude limitation into a running module and synchronously analyzing bidirectional dynamic response data, the system can realize online quantification of dynamic stability margin and collaborative compensation capability of the system, and further construct a high-confidence health trend atlas.

[0017] The system specifically comprises a main controller, a plurality of power modules, and the power modules comprise a collaborative processing unit and a boundary control unit controlled by the collaborative processing unit; wherein the main controller is used for:

[0018] S1, determining whether a current system running state meets a preset perturbation test safety condition; wherein the perturbation test safety condition comprises that the system is in a steady-state working state, a total load rate of the system is lower than a safety threshold, and no fault alarm occurs within a preset time length;

[0019] Specifically, the main controller continuously collects real-time running parameters reported by the collaborative processing units of the power modules during system running. The parameters include but are not limited to output voltages of the modules, output currents of the modules, output powers of the modules, input voltages of the modules, input currents of the modules, input powers of the modules, output voltages of the boundary control units, output currents of the boundary control units, output powers of the boundary control units, input voltages of the boundary control units, input currents of the boundary control units, input powers of the boundary control units, and the like. Output current And the module's internal status flags. The main controller performs three consecutive decision operations based on this real-time data.

[0020] The first judgment operation is to verify whether the system is in a steady-state operating state. The main controller sets a length of... The time window, the length of which is determined based on the system's main dynamic response time constant, can be set to, for example, 5 to 10 times the period corresponding to the system's closed-loop bandwidth. Within this time window, the main controller calculates key bus parameters, such as the system's total output voltage. Or total output current The variance of the change Total system output voltage The total system output current can be obtained by measuring the voltage at key load points. The sum of the output current of all power modules The steady-state criterion formula is as follows: ,in The steady-state threshold voltage or current is based on the allowable fluctuation range of normal system operation, for example, based on measured data from 200 sets of industrial sensors. The value is set to 0.5% of the rated voltage. If the variance is less than the square of the threshold, the system is considered to be in steady-state operation.

[0021] The second judgment operation is to verify whether the total system load rate is lower than the safety threshold. The main controller calculates the total output power of all online power modules at the current moment. and the rated total output power capacity of the system design. Compare. Total system load rate. From the formula Calculated. Safety threshold. This is a preset percentage value, such as 70%, set to ensure that the system has sufficient power margin to cope with possible transient processes during perturbation testing and to avoid triggering overall protection. If If so, then this condition is deemed satisfied.

[0022] The third judgment operation is to verify that there are no fault alarms within a preset time period. The main controller accesses its maintained system-level fault alarm log, which records fault events such as overvoltage, overcurrent, and overheating reported by the collaborative processing units of various modules, along with their timestamps. Preset time period Based on the system reliability maintenance strategy, for example, it can be set to 1 hour to ensure that the test is conducted in a relatively stable and healthy baseline environment. The main controller checks the current time. Before Are there any fault alarm records within the specified time period? If not, this condition is considered met.

[0023] All three judgment operations must pass before the main controller can finally confirm that the current system operating state meets the perturbation test safety conditions, and then allow the process to proceed to the next step S2. If any judgment fails, the main controller will not initiate the perturbation test and will continue to monitor the system state until the conditions are met.

[0024] For example, a system comprising four power modules with a total rated power It is 2000 watts. The main controller collects data within the time window. Bus voltage within milliseconds Sample, calculate its variance The steady-state threshold voltage Set to 0.05 volts, its square ,satisfy Therefore, the system is in a steady-state operation. Simultaneously, the main controller calculates the current total output power. Watts, total system load rate Below the safety threshold Furthermore, the fault alarm log shows that the most recent alarm occurred 2 hours ago, meeting the preset condition of no fault alarms within 1 hour. Therefore, the main controller determines that all the safety conditions for the perturbation test are met.

[0025] S2. If so, then select one of the multiple power modules as the current test module according to the predetermined perturbation test sequence.

[0026] Optionally, selecting one power module from a plurality of power modules as the current test module according to a predetermined perturbation test sequence includes:

[0027] Calculate the difference between the health potential index of each power module and the preset benchmark index, sort them from largest to smallest, and select the difference of the first sorted module as the index comparison value.

[0028] When the index comparison value is greater than the preset first threshold, the power modules are selected in turn as the current test module according to the preset numbering order of the multiple power modules.

[0029] When the index comparison value is greater than a preset second threshold and less than or equal to the first threshold, the time interval since the last perturbation test is obtained, and the power module is selected as the current test module based on the size of the time interval.

[0030] When the index comparison value is less than or equal to the second threshold, the health status assessment value of each power module is obtained, and the power module is selected as the current test module according to the magnitude of the health status assessment value.

[0031] Specifically, after step S1 confirms that the perturbation test safety conditions are met, the main controller immediately executes a predetermined perturbation test sequence to select the current module under test. The core logic of this sequence is to make decisions based on a quantitative comparison of the performance states of each power module. The main controller first reads the latest recorded health potential index of all online power modules from its internal storage. subscript Representing the The power module is numbered. Simultaneously, the main controller reads a preset reference index. This benchmark index represents the expected value of the power module's health potential index under brand-new or ideally calibrated conditions. It can be set based on factory calibration values ​​or the median of long-term system operation statistics. For each power module, the main controller calculates the difference between its health potential index and the benchmark index. After completing the calculations, the main controller processes all... The values ​​are sorted from largest to smallest, and the largest difference is selected as the first value in the sorted list, which is then defined as the exponential comparison value. .

[0032] Index Comparison Value The value will guide the main controller into three different selection logic branches. First threshold With the second threshold It is a preset key judgment threshold, and satisfies The two thresholds are set based on the classification of the degree of system performance discretization, for example, based on historical fault data analysis. Set to 0.15. Setting it to 0.05 means that when the performance degradation of a module deviates significantly from the baseline, a balanced testing strategy is adopted, while when the performance is generally good, a more refined health-based strategy is adopted.

[0033] In the first branch, when the exponent is compared to the value When this occurs, it indicates that at least one power module's health potential is significantly lower than the baseline, and the overall system state is uneven. At this point, the main controller adopts a simple round-robin mechanism, querying its internal records for the power module number that last performed the perturbation test. Then, select according to the preset numbering sequence of the power modules. The next numbered power module is then designated as the current module under test. If... If it is the last number, then the loop continues to the first number.

[0034] In the second branch, when the exponent comparison value satisfies At this point, it indicates that there are some differences in the states of each module, but they are not significant. At this time, the main controller introduces a time dimension for selection. It obtains the time interval between the last time each power module underwent a perturbation test. This data is recorded and maintained by the main controller after each test. The selection logic prioritizes the module with the longest test interval, and its selection is based on a formula. Implementation, i.e., selection The power module with the highest value is selected as the current module under test. This ensures that all modules are tested within a similar timescale.

[0035] In the third branch, when the exponent is compared to the value... At this point, it indicates that the health potential index of all power modules is close to or better than the benchmark, and the system is in a good and uniform state. At this time, the main controller switches to relying on the independent health status assessment value of each power module. The selection process involves an evaluation value typically calculated based on long-term factors such as module runtime, temperature history, and efficiency curves. The selection objective focuses on modules with higher potential risks, based on a formula... Implementation, namely, selecting health status assessment values. The smallest power supply module is selected as the current module under test. In this way, when the system is in good condition, test resources are tended to be allocated to modules with relatively weak historical evaluations, enabling preventative maintenance.

[0036] Finally, the main controller identifies the identifier of the selected power module as the current module under test and passes it to the subsequent step S3.

[0037] For example, the system has four power modules with preset reference indices. The main controller reads their current health potential indices as follows: , , , The difference was calculated. , , , After sorting, the maximum difference This value comes from module 2. A preset first threshold is set. Second threshold .because satisfy Therefore, the system proceeds to the third selection branch. The main controller reads the health status assessment values ​​of each module. , , , .according to According to the rule, module 3 with the smallest evaluation value is selected as the current module under test.

[0038] S3. Send a boundary perturbation command to the collaborative processing unit of the current module under test. The collaborative processing unit of the current module under test controls the boundary control unit of the current module under test to perform a performance boundary adjustment operation within a preset perturbation duration. The boundary perturbation command includes a voltage perturbation command or a current limit perturbation command.

[0039] Optionally, the performance boundary adjustment operation includes:

[0040] Send a boundary perturbation command to the collaborative processing unit of the currently tested module. The boundary perturbation command includes a voltage perturbation command or a current limit perturbation command.

[0041] The collaborative processing unit of the currently tested module receives the voltage perturbation command, controls the boundary control unit of the currently tested module to adjust the output voltage setpoint of the currently tested module, and restores it after a preset perturbation duration; or,

[0042] The collaborative processing unit of the currently tested module receives the current limit perturbation command, controls the boundary control unit of the currently tested module to reduce the output current protection threshold of the currently tested module, and recovers after the duration of the perturbation.

[0043] Specifically, after successfully selecting the current module under test in step S2, the main controller immediately generates a boundary perturbation command. This command is generated based on a predefined perturbation test mode, which can be fixed or dynamically selected based on system operating history. The boundary perturbation command contains a clear command type identifier, namely a voltage perturbation command or a current limit perturbation command. The command content also includes specific perturbation parameters, most importantly the perturbation amount. Duration of perturbation .

[0044] perturbation quantity The setting is based on the principle of both stimulating observable dynamic responses from the module and ensuring that no system-level faults are triggered under the safe conditions confirmed in step S1. For voltage perturbation commands, This is reflected in the voltage regulation amount. Its value can be set to the module's rated output voltage. A small percentage, such as one to three percent, chosen based on module loop stability design margin. At this point, the target voltage setpoint encoded in the instruction... From the formula The calculation shows that, among which This is the original output voltage setting of the module before the perturbation. For current limit perturbation commands, This is reflected in the reduction of the current protection threshold. It can be set to the module's rated output current. Five to ten percent, based on simulating a slight overload condition to test the current limiting response speed. At this point, the new current protection threshold encoded in the instruction... From the formula Calculations show that This is the original threshold.

[0045] Perturbation duration This is a key parameter; its setting must be greater than the main settling time of the control loop of the module under test, but much smaller than the system thermal time constant, typically ranging from tens to hundreds of milliseconds. This value is determined based on the module's power stage circuitry and control bandwidth, for example, through a step response test. It is set to be 2 to 3 times the time required for the module to reach a new steady state.

[0046] The main controller sends the encapsulated boundary perturbation command to the co-processing unit of the currently tested module via the system's internal communication bus. Upon receiving the command, the co-processing unit first parses the command type and parameters. If the parsing result is a voltage perturbation command, the co-processing unit immediately issues a command to the boundary control unit within the same module, requesting it to adjust the module's output voltage feedback loop reference value, i.e., the output voltage setpoint, from the current... Change to The boundary control unit, typically implemented by a digital controller or analog reference circuit, performs the change, forcing the module's power circuitry to adjust its output, resulting in a voltage step.

[0047] If the analysis result is a current limit perturbation command, the co-processing unit instructs the boundary control unit to adjust the overcurrent protection comparator threshold or digital current limit setting of the module, reducing it from... Reduce to This does not immediately change the module's output current, but once the load or system conditions cause the module's output current to approach this new threshold, the current limiting protection mechanism will intervene in advance, thereby testing the module's regulation and protection characteristics under boundary conditions.

[0048] Simultaneously with issuing the adjustment command, the collaborative processing unit starts a precise timer whose duration is equal to the duration of the received perturbation. When this timer overflows, the coprocessor immediately sends a recovery command to the boundary control unit, instructing it to restore the output voltage setpoint to its original value. Or restore the output current protection threshold to The entire performance boundary adjustment operation, from executing the adjustment to restoring the original state, is strictly limited to... Complete within the time window.

[0049] For example, the main controller selects module 2 as the current module under test and decides to apply a voltage perturbation. The rated output voltage of module 2... The current output voltage setting is 12 volts. The voltage is 12 volts. The voltage adjustment is calculated based on a 2% perturbation. Target voltage setpoint Duration of perturbation The module bandwidth is set to 100 milliseconds. The main controller generates and sends a voltage perturbation command containing these parameters. Upon receiving this command, the co-processing unit of module 2 adjusts the output voltage setpoint to 12.24 volts and starts a 100-millisecond timer. After the timer expires, the co-processing unit controls the boundary control unit to restore the setpoint to 12 volts.

[0050] Optionally, the main controller is further configured to:

[0051] Send a preset calibration perturbation command to any power module;

[0052] Receive the output voltage setting value or the output current protection threshold change measurement data reported by the corresponding power module to obtain the actual boundary change measurement data;

[0053] The measurement data of the change in the output voltage setting value or the output current protection threshold reported by the power module when it receives the calibration perturbation command at the initial time is obtained to obtain the reference boundary change measurement data.

[0054] Compare the difference between the actual boundary change measurement data and the reference boundary change measurement data;

[0055] If the difference value exceeds the preset tolerance limit, a perturbation calibration coefficient for the corresponding power module is generated.

[0056] The perturbation calibration coefficient is used to correct the boundary perturbation commands subsequently sent to the corresponding power module.

[0057] Specifically, in addition to executing the regular sequence of steps S1 to S3, the main controller can also proactively initiate a calibration process when the system meets the safety conditions for perturbation testing or within a specific maintenance cycle. The main controller selects any power module as the calibration target and sends a preset calibration perturbation command to its co-processing unit. This command has the same format as the boundary perturbation command in step S3, including a voltage or current limit perturbation type and a specific perturbation quantity. and the duration of the perturbation . Typically, a moderate, accurately measurable value is chosen, such as one percent of the rated voltage, based on the premise that good signal-to-noise ratio data can be obtained within the module's measurement accuracy range.

[0058] After receiving the calibration perturbation command, the collaborative processing unit of the calibration object module executes the same process as step S3: the control boundary control unit adjusts the output voltage setpoint or output current protection threshold, and... After recovery. Meanwhile, the module's internal measurement circuitry, typically a high-precision analog-to-digital converter, monitors and records in real-time the actual change in the setpoint executed by the boundary control unit. For voltage perturbations, it measures the actual change in the output voltage feedback reference point. For current limit perturbations, the measurement focuses on the actual change in the protection comparator threshold point. After the perturbation operation is completed, the collaborative processing unit reports the measured actual boundary change data to the main controller.

[0059] The main controller receives the reported actual boundary change measurement data. Then, it will query its own non-volatile memory for the measurement data reported by the calibration target module at the initial moment, such as when the module is first powered on for calibration or when it leaves the factory, when it receives the same calibration perturbation command. This data is defined as the reference boundary change measurement data. Initial baseline data is acquired during system construction or module replacement by performing the same calibration procedure once and permanently storing the results.

[0060] Next, the main controller compares the actual measured data with the reference data. The difference value... Through formula The difference was calculated. It needs to be consistent with a preset tolerance limit. Compare. Tolerance limits. The setting is based on the long-term drift range allowed by the analog and digital control circuits of the power module. For example, based on the temperature drift and time drift specifications of key components such as the reference voltage source and resistors, it is set to two percent of the rated perturbation.

[0061] If the calculated difference value Exceeding the tolerance limit ,Right now This indicates that the module's actual response to the same perturbation command has deviated significantly from its initial state. To compensate for this deviation, the main controller generates a perturbation calibration coefficient for the power supply module. The coefficient is calculated based on the proportional relationship between the benchmark and the actual measured value, and the formula is as follows: . It is a dimensionless correction factor with a value close to 1. Less than ,but Conversely, it is less than 1.

[0062] Generate perturbation calibration coefficients The main controller then stores it in association with the module's identifier. Subsequently, whenever any boundary perturbation command needs to be sent to this specific power module, whether it's the test command from step S3 or a subsequent calibration command, the main controller applies this coefficient for correction when generating the command. The correction method involves modifying the original perturbation amount encoded in the command. Multiply by the perturbation calibration factor to obtain the corrected perturbation. The main controller uses This process constructs and sends the final instruction data packet. This correction process ensures that the actual boundary changes expected to occur after the issued instruction has undergone module hardware offset compensation are consistent with the original intention of the instruction, thereby guaranteeing the consistency of the perturbation excitation on which subsequent health status assessments depend.

[0063] For example, the main controller initiates voltage calibration for power module 1. A preset calibration perturbation amount... To increase the voltage by 0.20 volts. After module 1 executes, it reports the actual boundary change measurement data. The main controller retrieved the reference boundary change measurement data for this module. Calculate the difference value. Preset tolerance limit .because equal The judgment condition is "exceeding", that is... The calibration is only triggered at this point, and the conditions are not met, so no perturbation calibration coefficients are generated. In another scenario... ,but The conditions are met. The main controller generates perturbation calibration coefficients. Subsequently, if a test command with a perturbation of 0.10 volts needs to be sent to module 1, the main controller will correct it to... Send again.

[0064] S4. During the duration of the perturbation, obtain first type of response data from the collaborative processing unit of the current tested module, and analyze and calculate the dynamic stability margin index of the current tested module based on the first type of response data.

[0065] Optionally, the analysis and calculation of the dynamic stability margin index of the currently tested module includes:

[0066] The pulse width modulation signal duty cycle of the current module under test is obtained from the collaborative processing unit of the current module under test, the trajectory data of the change of the pulse width modulation signal duty cycle during the duration of the perturbation, and the stabilization time data of the output filter inductor current after the output voltage setting value or the output current protection threshold is recovered, as the first type of response data.

[0067] From the duty cycle change trajectory data of the first type of response data, the final change in duty cycle adjustment and the adjustment speed at which the final change is reached are extracted as key features;

[0068] Establish a proportional relationship model between the key features and the output voltage setting value or the output current protection threshold;

[0069] The settling time data is used as a metric for the system's damping characteristics.

[0070] By combining the aforementioned proportional relationship model with the aforementioned metric, a dynamic stability margin index is calculated.

[0071] Specifically, at the same time that the performance boundary adjustment operation initiated in step S3 begins execution of the current module under test, the main controller starts synchronous acquisition of the first type of response data. The main controller sends a data stream request command to the co-processing unit of the current module under test, and the co-processing unit of that module then begins to record and buffer two key internal real-time signals at a high sampling rate, for example, no less than 10 times the switching frequency.

[0072] The first signal is the duty cycle of the pulse width modulation signal, denoted as... This signal is the core output of the module's internal control loop, directly driving the power switching devices; its value continuously varies between 0 and 1. The co-processing unit records the time from the start of perturbation adjustment. When the perturbation is recovered, the system re-enters steady state. Throughout the entire time period The continuous values ​​form the duty cycle change trajectory data.

[0073] The second signal is the output filter inductor current, denoted as... The co-processing unit acquires the real-time value of this current through a built-in current sensor and analog-to-digital converter, paying particular attention to the process of the current value stabilizing again after the disturbance duration ends and the boundary conditions return to normal. The co-processing unit calculates and records the stabilization time data of the output filter inductor current after the output voltage setpoint or output current protection threshold is restored. . Defined from the recovery time Beginning, to Enter and remain at its final steady-state value Centered on, with a width of The time elapsed within the zone. It is usually set to one to two percent of the rated current, based on the system's noise level and measurement accuracy.

[0074] After the perturbation test cycle ends, the collaborative processing unit of the currently tested module will provide complete duty cycle change trajectory data and settling time data. Package it and upload it to the main controller as the first type of response data.

[0075] After receiving the first type of response data, the main controller begins analysis and calculation. This starts with the duty cycle change trajectory data. Two key features were extracted. The first key feature is the final change in the duty cycle adjustment. Through formula Calculation, where It is the duty cycle just before the perturbation occurs. It is the duty cycle at which the system reaches a new steady state after the perturbation is applied. The second key feature is the rate of adjustment to reach the final change. This speed can be calculated by determining the time required for the duty cycle to change from its initial value to 90% of its final change. Quantified by the reciprocal, that is Its dimensions are .

[0076] Subsequently, the main controller establishes a proportional relationship model between these key characteristics and the perturbation applied in step S3. This model describes the static response gain of the module control loop to boundary changes. For voltage perturbations, the model is as follows: For the current limit perturbation, the model is as follows: This proportionality coefficient This model comprehensively reflects the current state of the module's power stage parameters and control parameters, and their values ​​should be within a reasonable range. The model is based on the small-signal model principle of switching power supply systems.

[0077] At the same time, the main controller will stabilize the time data. As a measure of the system's damping characteristics. A shorter one. This usually means the system has stronger damping and a faster response. The settling time is related to the system damping ratio. and natural frequency Related, can be found through formula Perform correlation estimation, where It is the allowable error range percentage, for example, corresponding to the previous... .

[0078] Finally, the main controller, combining the proportional relationship model and damping metric, calculates the dynamic stability margin index. This indicator needs to comprehensively reflect the module's static adjustment capability and dynamic stability speed. An example calculation formula is as follows: ,in It is a normalization coefficient used to ensure that the result falls within a standard range of 0 to 1. Its setting is based on statistical analysis of a large amount of historical test data. In the formula... The aforementioned proportionality coefficient has the following dimensions: or , Dimensions are , Dimensions are Through coefficients Perform dimensional normalization, and finally The value is dimensionless. The calculated value is... The higher the value, the more dynamic stability margin the tested module has when dealing with boundary disturbances.

[0079] For example, the module under test currently performs a voltage perturbation, the perturbation amount is... The duty cycle change trajectory data acquired by the main controller shows that... Under the new steady state Therefore Calculate the proportionality coefficient. The time taken for the duty cycle to change from 0.50 to 0.518 (i.e., 0.9 * 0.02 + 0.50) is... Adjust speed Stable time data The natural frequency of the system design is known. The damping ratio can be inversely calculated from the settling time formula. Set the normalization coefficient. Calculate the dynamic stability margin index .

[0080] S5. During the duration of the perturbation, obtain second type of response data from the collaborative processing unit of at least one other power module, and analyze and calculate the collaborative compensation capability index of the system based on the second type of response data.

[0081] Optionally, the collaborative compensation capability index of the analysis and calculation system includes:

[0082] The output current increment data or output voltage adjustment data of the corresponding power module during the duration of the perturbation are obtained from the cooperative processing unit of at least one other power module as the second type of response data.

[0083] Based on the output current increment data or output voltage adjustment data in the second type of response data, the power modules that have undergone compensatory adjustments are selected to obtain the collaborative compensation group;

[0084] Calculate the ratio of the total compensation output change of the collaborative compensation group to the bus parameter change during the duration of the perturbation to obtain the system static compensation gain index;

[0085] The response delay data and smoothness data of the compensation action of the collaborative compensation group are obtained and analyzed to obtain the dynamic compensation quality index of the system.

[0086] By combining the static compensation gain index and the dynamic compensation quality index of the system, the collaborative compensation capability index is calculated.

[0087] Specifically, when the currently tested module performs a performance boundary adjustment operation in step S3, the power balance of the entire system is briefly disrupted. To assess the coordinated response capability of other power modules within the system to this disturbance, the main controller, simultaneously initiating the perturbation command or within a very short communication delay thereafter, sends a data monitoring command to all non-tested power modules, i.e., the collaborative processing units of the other power modules. The command requires these modules to report the dynamic changes of their key output parameters within the perturbation duration window.

[0088] Each of the other power modules' collaborative processing units continuously monitors and records specified data based on the instruction type. If the main controller focuses on evaluating current sharing compensation capabilities, the instruction requires reporting the output current increment data. This data represents the module's output current relative to its steady-state value before the disturbance occurred. The real-time difference. If the focus is on evaluating the voltage regulation compensation capability, the instruction requires reporting the output voltage regulation data. This refers to the real-time deviation of the output voltage from its original set value. (Subscript) Representing the Other power modules. This data is sampled at a certain time resolution and reported to the main controller as Type II response data after the perturbation test is completed.

[0089] After collecting all reported Type II response data, the main controller begins screening power modules that have undergone compensatory adjustments. The screening is based on whether the module's output change has a directional tendency to compensate for the disturbance in the module under test. For example, if the module under test performs a perturbation that increases its output voltage, it will theoretically tend to reduce its output current. To maintain bus voltage stability, other healthy modules should increase their output current. Therefore, the main controller calculates the steady-state average of the output change of each other module during the perturbation. ,in Represents current or voltage Set a judgment threshold. This threshold is set based on system noise and the minimum effective compensation, for example, 0.5% of the rated current. If The absolute value is greater than If the sign of the module matches the expected compensation direction, then the module is determined to have undergone compensatory adjustment and is included in the collaborative compensation group. The collaborative compensation group is a dynamically selected set of modules, denoted as set . .

[0090] Next, the main controller calculates the system's static compensation gain index. This index quantifies the static proportional relationship between the total compensation output of the coordinated compensation group and the initial disturbance. First, the change in the total compensation output of the coordinated compensation group is calculated. Simultaneously, the main controller obtains the changes in bus parameters during the duration of the disturbance from the system bus monitoring unit, such as the changes in bus voltage. or the change in total load current The system static compensation gain index is given by the formula. The calculation shows that, among which This represents the corresponding bus parameters. Ideally, this ratio should be close to 1, indicating that the disturbance has been fully compensated. and Both are either current or both are voltage.

[0091] Then, the main controller analyzes the system's dynamic compensation quality indicators. This involves timing analysis of the response process of the coordinated compensation group. The main controller extracts two dynamic features from the second type of response data. The first is the response delay data. Defined as the time from the start of the perturbation Changes in module output For the first time, it exceeded its steady-state average value. The average response delay of the collaborative compensation group is 20% of the time elapsed. It was calculated, of which The first is the number of modules in the collaborative compensation group. The second is the smoothness data of the compensation action, calculated by taking the root mean square value of the first derivative (i.e., the rate of change) of the trajectory of the output change of each module. To assess this, an excessively high rate of change indicates abrupt compensation actions and may trigger secondary oscillations. The overall smoothness of the coordinated compensation group... It can be defined as ,in This is a small constant added to prevent division by zero. System dynamic compensation quality index. Then it can be derived from the formula Comprehensive calculation, It is a normalization coefficient used to adjust the result to a suitable numerical range. Its setting is based on the typical time constant and rate of change level of the system's dynamic response.

[0092] Finally, the main controller combines the system's static compensation gain index and the system's dynamic compensation quality index to calculate the final collaborative compensation capability index. A feasible comprehensive formula is as follows: ,in It is another comprehensive normalization coefficient that makes Ultimately, the score falls within a standardized range, such as 0 to 100. Higher scores... The value indicates that when the system responds to disturbances in local modules, the other modules can compensate quickly, smoothly, and sufficiently, demonstrating good system coordination.

[0093] For example, the module under test performs an output voltage up-adjustment perturbation, causing a momentary increase in the bus voltage. Three other modules in the system reported output current increment data, with steady-state average values ​​of [data missing]. , , Set a judgment threshold. Both Module 1 and Module 2 change in a positive direction (increasing output to reduce bus voltage), but the change in Module 2 is 0.1A, which is less than 0.3A. Therefore, only Module 1 is selected for the collaborative compensation group. Total compensation output change System static compensation gain index The dimension here is admittance, which can be normalized to a dimensionless value by normalizing it with the system characteristic admittance (such as 1 / load impedance). However, for clarity of example, the original value is used for now. Assume the response delay of module 1 is... The root mean square rate of change of its current ,Pick , smoothness .set up ,but The dimensions of this value are The value is relatively large and requires subsequent normalization. Let... ,but .

[0094] S6. Integrate the dynamic stability margin index and the collaborative compensation capability index to generate the health potential index of the currently tested module.

[0095] Optionally, generating the health potential index of the currently tested module includes:

[0096] The dynamic stability margin index and the collaborative compensation capability index are normalized to obtain standardized margin scores and capability scores, respectively.

[0097] The margin score is weighted according to the preset importance weight of the currently tested module in the system.

[0098] The weighted margin score and the capability score are combined according to a preset ratio to generate a comprehensive score;

[0099] The comprehensive score is mapped to a health potential index.

[0100] Specifically, the dynamic stability margin index of the current tested module is calculated in steps S4 and S5 respectively. Cooperative compensation capability index of the system Next, the main controller executes an indicator fusion process to generate a comprehensive health potential index. Because... and These two indicators may have different dimensions and numerical ranges, making direct fusion incomparable. Therefore, the main controller first normalizes them, converting them into standardized dimensionless scores.

[0101] Normalization requires a reference baseline. The main controller reads the theoretical maximum value of the dynamic stability margin index from the system's historical database or a preset configuration table. and minimum value And the theoretical maximum value of the collaborative compensation capability index. and minimum value These extreme values ​​are determined based on the boundary between system design specifications and long-term operational statistics. For example, It can be set to 90% of the statistical upper limit obtained from testing all modules under brand-new and ideal calibration conditions. Subsequently, margin scores are calculated separately. With ability score :

[0102] ,

[0103] ,

[0104] Calculated and Constrained within the range of 0 to 1, a higher value indicates better performance on that single indicator.

[0105] Next, the main controller determines the importance weight of the currently tested module within the system. The margin scores are weighted. Importance weights. This is a coefficient greater than 0, and its setting is based on the load proportion borne by the module, whether it is a critical backup module, or its position in the topology. For example, a module bearing core load might be assigned a weight of 1.2, while a module with general load might be assigned a weight of 1.0. The weighted margin score is then calculated. From the formula The calculation yielded the result.

[0106] Subsequently, the main controller will calculate the weighted margin score. With ability score The indicators are synthesized according to a preset ratio. This preset ratio reflects the relative importance of the two types of indicators in the final evaluation, and is determined by weighting coefficients. and It means that, among them It is a preset value between 0 and 1, for example, 0.6 is set based on expert experience or historical fault correlation analysis. Overall Score Calculated using the linear weighted composition formula:

[0107] ,

[0108] Overall score This comprehensively reflects the combined effect of the module's own stability margin and the system's support capability for it.

[0109] Finally, the main controller will calculate the overall score. Mapped to the final health potential index The mapping relationship can be linear or non-linear. A simple linear mapping formula is: ,in This is the baseline value for the health potential index, for example, set to 60. It is a proportionality coefficient used to map the composite score from 0 to 1 to a meaningful exponential range, such as 0 to 40, thereby... It falls between 60 and 100. (Proportionality coefficient) Compared with the benchmark value The design is based on the principle that the numerical range of the health potential index can intuitively distinguish between health status levels such as "good," "attention," and "warning." The final generated... The value will be stored and associated with the identifier of the currently tested module for subsequent system health management and decision-making.

[0110] For example, suppose that after one test, the main controller obtains a dynamic stability margin index. Collaborative compensation capability indicators The system's default settings , ; , Calculate the margin score. Ability score Importance weight of the currently tested module. After weighting Preset synthesis ratio Then the overall score .set up , Then the health potential index .

[0111] S7. Construct and update the system health potential map based on the health potential index of each power module; when load scheduling is required, schedule the power modules to unload the load based on the health potential index in the system health potential map; when standby modules need to be activated, select the standby power modules to connect to the system based on the health potential index in the system health potential map.

[0112] Specifically, the main controller maintains a dynamically updated data structure called the system health potential map. This map uses the unique identifier of each power module as an index to store its latest health potential index. and the latest update timestamp of the index. Whenever step S6 completes the calculation and generation of the health potential index for any power module, the main controller immediately uses the new index. and the time when the calculation is completed Update the data of the corresponding index entries in the system's health potential graph. The essence of building the graph is to establish a mapping relationship between module identifiers and real-time health potential status, and its data structure can be a hash table or a database table.

[0113] The update logic for the system health potential map not only covers index values ​​but also includes tracking historical trends. The main controller can be configured to retain only the most recent N valid values. The value and its timestamp, N, are set based on the system's requirements for historical data backtracking depth; for example, it is set to 30 for monthly health analysis. When an entry in a module of the graph is overwritten by new data, the old data is moved to the history stack, with the latest data always at the top. This design ensures that the graph not only reflects the current state but also provides a data foundation for subsequent trend analysis, such as step S8.

[0114] When load scheduling is required during system operation, such as when the total load power increases and needs to be shared by online modules, or when a module needs to operate at reduced capacity, the main controller makes decisions based on the system health potential map. The core objective of load scheduling is to prioritize ensuring that modules with better health and greater potential can handle more load, thereby optimizing the overall system reliability. The main controller first extracts the latest health potential index of all currently online and non-standby power modules from the system health potential map, forming a set. Then, the main controller follows... The set is sorted from lowest to highest value. The load offloading scheduling decision is: select... The power module with the lowest value is selected for load offloading. An offloading command is sent to the module's co-processing unit, instructing it to gradually reduce its output power setting by a preset step size, such as five percent of its rated power, until the load demand is met or its safe operating limit is reached. This selection logic is based on the fact that modules with lower health potential indices have a relatively higher risk of future performance degradation or failure; therefore, prioritizing load reduction for these modules helps extend their lifespan and reduce system operational risks.

[0115] On the other hand, when the system needs to activate a backup module, such as when there are insufficient online modules or when planned maintenance needs to be performed, the main controller selects from all power modules currently in standby mode based on the system health potential map. The main controller extracts the latest health potential index of all power modules marked as standby mode from the map, forming a set. Conversely, the logic for enabling standby modules is to select the standby module with the highest health potential index. The main controller follows... Values ​​in descending order Sort the set and select the first-ranked item. The standby power module with the highest health potential index is selected for connection. Once selected, the main controller sends wake-up and grid connection commands to the module's co-processing unit, controlling its boundary control unit to execute a soft-start process, smoothly synchronizing its output with the system bus before putting it into operation. This selection logic is based on the fact that the standby module with the highest health potential index is considered to be in its best current state, providing the most reliable power output after being connected to the system, and possessing a stronger potential to cope with future disturbances.

[0116] For example, a system comprises five power modules, whose identifiers and the latest health potential index recorded in the graph are as follows: Module A, (Online), Module B, (Online), Module C, (Online), Module D, (Standby), Module E, (Standby). When the system detects a decrease in total load and needs to unload the load of a module, the main controller determines the load based on the online module set. The numbers, sorted from lowest to highest, are 85, 88, and 92. According to the scheduling rules, [the following is selected]. Module B, corresponding to a minimum value of 85, performs load offloading. If the system subsequently needs to activate a standby module due to increased load, the main controller will determine the standby module set. Sort by highest to lowest as 95 and 90. Based on the activation rules, select... The highest value of 95 corresponds to module D accessing the system.

[0117] S8. Based on the system health potential map, obtain the time-series data of the health potential index of any power module and calculate the time-series change rate of the potential index; obtain the health status assessment value of the corresponding power module at the corresponding time and calculate the time-series change rate of the status assessment; based on the time-series change rate of the potential index and the time-series change rate of the status assessment, calculate the health trend correction coefficient of the corresponding power module; use the health trend correction coefficient to correct the health status assessment value of the corresponding power module.

[0118] Specifically, to perform trend correction analysis, the main controller first extracts the health potential index sequence of the target power module at multiple consecutive time points from the historical records of the system health potential map. These time points typically correspond to the completion times of each perturbation test, forming a time series. and its corresponding health potential index sequence This is the time-series data for the health potential index. (Sequence length) The setting is based on the trend stability required for the analysis, such as setting it to no less than 10 sets of data based on statistical significance requirements.

[0119] After acquiring the time-series data, the main controller calculates the time-series change rate of the potential index. This rate of change reflects the trend of the health potential index over time and is calculated using a linear regression method. The main controller uses time... As the independent variable, with Fit a straight line to the dependent variable. Among them, the slope This refers to the time-series rate of change of the potential index. The calculation process uses the least squares method, and the formula is as follows: . The numerical sign indicates the trend direction: a positive value indicates an upward trend, and a negative value indicates a downward trend.

[0120] Synchronously, the main controller retrieves the sequence of the same power module at exactly the same time point from the independently maintained health status assessment history. Health status assessment value sequence Health status assessment values ​​are typically calculated based on long-term accumulated parameters such as runtime, temperature, and efficiency. The update cycle may differ from that of perturbation testing, but the main controller will extract values ​​related to these parameters. The timestamps accurately correspond to the evaluation values, ensuring that the data is aligned in the time dimension.

[0121] Based on the aligned health status assessment time-series data, the main controller calculates the time-series change rate of the status assessment using the same method. Fitted straight line Its slope This refers to the rate of change of the state assessment over time. The calculation formula is as follows: .

[0122] Two rates of change were obtained. and Then, the main controller calculates the health trend correction coefficient based on them. This coefficient aims to quantify the consistency between the changing trends reflected in short-term dynamic tests and those reflected in long-term cumulative assessments, and to correct for long-term assessments. The calculation logic is as follows: if... and Same symbols and If the value exceeds a set trend sensitivity threshold, it indicates that the dynamic test has captured a more significant or faster trend change, requiring adjustments to the long-term evaluation value accordingly. A specific calculation formula is as follows: .in, It is a preset correction strength coefficient, between 0 and 1, which is set based on the correlation strength between two trends in historical data, for example, 0.3 based on regression analysis. yes The sign function ensures that the correction direction is consistent with the long-term assessment trend. This is a reference rate of change used to normalize the difference, typically set as the absolute value of the average rate of change of the health status assessment value during the nominal lifespan, where it decreases linearly. This formula makes... It becomes a dimensionless multiplicative correction factor.

[0123] Finally, the main controller utilizes the health trend correction coefficient. Correct the current health status assessment value of the corresponding power module. The corrected formula is as follows: Revised health status assessment values This will replace the original value for subsequent system management decisions, such as module selection or load scheduling reference in step S2. This modification allows health status assessments based on long-term static parameters to incorporate the latest performance change trends revealed by short-term dynamic perturbation tests, enabling more timely and accurate assessments of module health status.

[0124] For example, for power module X, the main controller extracts the timestamps (in days, from 1 to 10) of its most recent 10 perturbation tests and the corresponding health potential index: The sequence is {95,94,93,92,91,90,89,88,87,86}. The time-series change rate of the potential index was calculated. Simultaneously, obtain health status assessment values ​​at the same 10 time points: The sequence is {92, 91.8, 91.6, 91.4, 91.2, 91.0, 90.8, 90.6, 90.4, 90.2}. The calculated time-series change rate of the state assessment is... Set the correction strength coefficient. Reference rate of change Calculate the health trend correction factor. Since the correction factor should generally be kept within a positive range to avoid contradictions in physical meaning, a lower limit can be set for it, for example, 0.5. If the lower limit is set to 0.5, then take... Assuming the current health status assessment value... After correction .

[0125] Based on the same inventive concept, such as Figure 2 As shown, the present invention also provides a modular power management method for embedded health status assessment, the method comprising:

[0126] The main controller determines whether the current system operating state meets the preset perturbation test safety conditions; wherein, the perturbation test safety conditions include the system being in a steady-state operating state, the total system load rate being lower than the safety threshold, and no fault alarms within a preset time period;

[0127] If so, the main controller selects one of the multiple power modules as the current test module according to the predetermined perturbation test sequence.

[0128] The main controller sends a boundary perturbation command to the collaborative processing unit of the currently tested module. The collaborative processing unit of the currently tested module controls the boundary control unit of the currently tested module to perform a performance boundary adjustment operation within a preset perturbation duration. The boundary perturbation command includes a voltage perturbation command or a current limit perturbation command.

[0129] During the duration of the perturbation, the main controller obtains first type of response data from the collaborative processing unit of the current module under test, and analyzes and calculates the dynamic stability margin index of the current module under test based on the first type of response data.

[0130] During the duration of the perturbation, the main controller obtains second-type response data from the cooperative processing unit of at least one other power module, and analyzes and calculates the cooperative compensation capability index of the system based on the second-type response data.

[0131] The main controller integrates the dynamic stability margin index and the collaborative compensation capability index to generate the health potential index of the currently tested module.

[0132] It should be noted that the electrical connections between the various units described above do not necessarily represent direct or indirect connections. Any method of indirect connection is applicable to the embodiments of the present invention as long as it achieves the purpose of the present invention. The above description is merely an exemplary embodiment of the present invention and should not be construed as limiting the scope of the present invention.

[0133] All equivalent changes and modifications made in accordance with the teachings of this invention are still within the scope of this invention. Those skilled in the art will readily conceive of other embodiments of this invention upon considering the specification and the disclosure of practical truth. This application is intended to cover any variations, uses, or adaptations of this invention that follow the general principles of this invention and include common knowledge or conventional techniques in the art not described herein.

Claims

1. A modular power management system for embedded health status assessment, characterized in that, The system includes: a main controller and multiple power modules, each power module including a collaborative processing unit and a boundary control unit controlled by the collaborative processing unit; wherein, the main controller is used for: Determine whether the current system operating status meets the preset perturbation test safety conditions; wherein, the perturbation test safety conditions include the system being in a steady-state operating state, the total system load rate being lower than the safety threshold, and no fault alarms within a preset time period; If so, then according to the predetermined perturbation test sequence, one of the multiple power modules is selected as the current test module; this includes: calculating the difference between the health potential index of each power module and the preset benchmark index, arranging them from largest to smallest, and selecting the difference ranked first as the index comparison value; when the index comparison value is greater than a preset first threshold, power modules are selected in turn as the current test module according to the preset numbering order of the multiple power modules; when the index comparison value is greater than a preset second threshold and less than or equal to the first threshold, the time interval since the last perturbation test is obtained, and the power module is selected as the current test module based on the size of the time interval; when the index comparison value is less than or equal to the second threshold, the health status assessment value of each power module is obtained, and the power module is selected as the current test module based on the size of the health status assessment value. A boundary perturbation command is sent to the collaborative processing unit of the currently tested module. The collaborative processing unit of the currently tested module controls the boundary control unit of the currently tested module to perform a performance boundary adjustment operation within a preset perturbation duration. The boundary perturbation command includes a voltage perturbation command or a current limit perturbation command. This includes: sending a boundary perturbation command to the collaborative processing unit of the currently tested module, the boundary perturbation command including a voltage perturbation command or a current limit perturbation command; the collaborative processing unit of the currently tested module receiving the voltage perturbation command, controlling the boundary control unit of the currently tested module to adjust the output voltage setting value of the currently tested module, and restoring after the preset perturbation duration; or, the collaborative processing unit of the currently tested module receiving the current limit perturbation command, controlling the boundary control unit of the currently tested module to reduce the output current protection threshold of the currently tested module, and restoring after the perturbation duration. During the duration of the perturbation, first-type response data is obtained from the collaborative processing unit of the current module under test, and based on the first-type response data, the dynamic stability margin index of the current module under test is analyzed and calculated. This includes: obtaining from the collaborative processing unit of the current module under test the trajectory data of the change of the duty cycle of the pulse width modulation signal of the current module under test during the duration of the perturbation, and the stabilization time data of the output filter inductor current after recovery from the output voltage setpoint or the output current protection threshold, as the first-type response data; extracting the final change in duty cycle adjustment and the adjustment speed to reach the final change from the duty cycle change trajectory data of the first-type response data as key features; establishing a proportional relationship model between the key features and the output voltage setpoint or the output current protection threshold; using the stabilization time data as a metric for the system damping characteristics; and combining the proportional relationship model and the metric to calculate the dynamic stability margin index. During the duration of the disturbance, second-type response data is obtained from the collaborative processing unit of at least one other power module, and the collaborative compensation capability index of the system is analyzed and calculated based on the second-type response data. This includes: obtaining output current increment data or output voltage adjustment data of the corresponding power module during the duration of the disturbance from the collaborative processing unit of at least one other power module, as the second-type response data; filtering out power modules that have undergone compensatory adjustments based on the output current increment data or output voltage adjustment data in the second-type response data to obtain a collaborative compensation group; calculating the ratio of the total compensation output change of the collaborative compensation group to the bus parameter change during the duration of the disturbance to obtain a system static compensation gain index; obtaining and analyzing the response delay data and smoothness data of the compensation action of the collaborative compensation group to obtain a system dynamic compensation quality index; and combining the system static compensation gain index and the system dynamic compensation quality index to calculate the collaborative compensation capability index. By integrating the dynamic stability margin index and the collaborative compensation capability index, a health potential index for the currently tested module is generated.

2. The modular power management system for embedded health status assessment according to claim 1, characterized in that, The generation of the health potential index of the currently tested module includes: The dynamic stability margin index and the collaborative compensation capability index are normalized to obtain standardized margin scores and capability scores, respectively. The margin score is weighted according to the preset importance weight of the currently tested module in the system. The weighted margin score and the capability score are combined according to a preset ratio to generate a comprehensive score; The comprehensive score is mapped to a health potential index.

3. The modular power management system for embedded health status assessment according to claim 2, characterized in that, The main controller is also used for: Based on the health potential index of each power module, construct and update the system health potential map; When load scheduling is required, the power modules are scheduled to unload the load based on the health potential index in the system health potential map. When a standby module needs to be activated, the standby power module is selected to be connected to the system based on the health potential index in the system health potential map.

4. The modular power management system for embedded health status assessment according to claim 1, characterized in that, The main controller is also used for: Send a preset calibration perturbation command to any power module; Receive the output voltage setting value or the output current protection threshold change measurement data reported by the corresponding power module to obtain the actual boundary change measurement data; The measurement data of the change in the output voltage setting value or the output current protection threshold reported by the power module when it receives the calibration perturbation command at the initial time is obtained to obtain the reference boundary change measurement data. Compare the difference between the actual boundary change measurement data and the reference boundary change measurement data; If the difference value exceeds the preset tolerance limit, a perturbation calibration coefficient for the corresponding power module is generated. The perturbation calibration coefficient is used to correct the boundary perturbation commands subsequently sent to the corresponding power module.

5. A modular power management system for embedded health status assessment according to claim 3, characterized in that, The main controller is also used for: Based on the system health potential map, time-series data of the health potential index of any power module are obtained, and the time-series change rate of the potential index is calculated. Obtain the health status assessment value of the corresponding power module at the corresponding time, and calculate the time-series change rate of the status assessment. Based on the time-series change rate of the potential index and the time-series change rate of the state assessment, the health trend correction coefficient of the corresponding power module is calculated. The health status assessment value of the corresponding power module is corrected using the health trend correction coefficient.

6. A modular power management method for embedded health status assessment, characterized in that, The method includes: The main controller determines whether the current system operating state meets the preset perturbation test safety conditions; wherein, the perturbation test safety conditions include the system being in a steady-state operating state, the total system load rate being lower than the safety threshold, and no fault alarms within a preset time period; If so, the main controller selects one of multiple power modules as the current test module according to a predetermined perturbation test sequence; this includes: calculating the difference between the health potential index of each power module and a preset benchmark index, arranging them from largest to smallest, and selecting the difference ranked first as the index comparison value; when the index comparison value is greater than a preset first threshold, selecting power modules as the current test module in turn according to the preset numbering order of the multiple power modules; when the index comparison value is greater than a preset second threshold and less than or equal to the first threshold, obtaining the time interval since the last perturbation test, and selecting a power module as the current test module based on the size of the time interval; when the index comparison value is less than or equal to the second threshold, obtaining the health status assessment value of each power module, and selecting a power module as the current test module based on the size of the health status assessment value. The main controller sends a boundary perturbation command to the co-processing unit of the currently tested module. The co-processing unit of the currently tested module controls the boundary control unit of the currently tested module to perform a performance boundary adjustment operation within a preset perturbation duration. The boundary perturbation command includes a voltage perturbation command or a current limit perturbation command. This includes: sending a boundary perturbation command to the co-processing unit of the currently tested module, the boundary perturbation command including a voltage perturbation command or a current limit perturbation command; the co-processing unit of the currently tested module receiving the voltage perturbation command, controlling the boundary control unit of the currently tested module to adjust the output voltage setting value of the currently tested module, and restoring after the preset perturbation duration; or, the co-processing unit of the currently tested module receiving the current limit perturbation command, controlling the boundary control unit of the currently tested module to reduce the output current protection threshold of the currently tested module, and restoring after the perturbation duration. During the duration of the perturbation, the main controller acquires first-type response data from the collaborative processing unit of the currently tested module, and analyzes and calculates the dynamic stability margin index of the currently tested module based on the first-type response data. This includes: acquiring, from the collaborative processing unit of the currently tested module, the trajectory data of the change of the duty cycle of the pulse width modulation signal of the currently tested module during the duration of the perturbation, and the stabilization time data of the output filter inductor current after recovery from the output voltage setpoint or the output current protection threshold, as the first-type response data; extracting the final change in duty cycle adjustment and the adjustment speed to reach the final change from the duty cycle change trajectory data of the first-type response data as key features; establishing a proportional relationship model between the key features and the output voltage setpoint or the output current protection threshold; using the stabilization time data as a metric for the system damping characteristics; and combining the proportional relationship model and the metric to calculate the dynamic stability margin index. During the duration of the disturbance, the main controller acquires second-type response data from the collaborative processing unit of at least one other power module, and analyzes and calculates the system's collaborative compensation capability index based on the second-type response data. This includes: acquiring, from the collaborative processing unit of at least one other power module, the output current increment data or output voltage adjustment data of the corresponding power module during the duration of the disturbance, as the second-type response data; filtering out power modules that have undergone compensatory adjustments based on the output current increment data or output voltage adjustment data in the second-type response data, thus obtaining a collaborative compensation group; calculating the ratio of the total compensation output change of the collaborative compensation group to the bus parameter change during the duration of the disturbance, thus obtaining a system static compensation gain index; acquiring and analyzing the response delay data and the smoothness data of the compensation action of the collaborative compensation group, thus obtaining a system dynamic compensation quality index; and combining the system static compensation gain index and the system dynamic compensation quality index to calculate the collaborative compensation capability index. The main controller integrates the dynamic stability margin index and the collaborative compensation capability index to generate the health potential index of the currently tested module.

Images