Parallel current sharing control method and system for multi-module power supply

Abstract

The application provides a multi-module power supply parallel current sharing control method and system, and particularly relates to the technical field of power electronics. The method comprises the following steps: collecting output current time sequence data of a plurality of parallel power supply modules, calculating current sharing residual errors of each parallel power supply module at different sampling moments; identifying a plurality of historical load events and constructing a plurality of residual error trajectories of each parallel power supply module, constructing a plurality of residual error slot occupation trajectory of each parallel power supply module according to the plurality of residual error trajectories, constructing an average residual error trajectory of each sampling moment in each type of historical load event, extracting slot occupation distribution data of each parallel power supply module at each sampling moment, constructing an identity slot trajectory of each parallel power supply module and generating a current sharing compensation trajectory of each parallel power supply module, and using the current sharing compensation trajectory for current sharing control of the plurality of parallel power supply modules. The application can improve the current sharing control pertinence and stability of the plurality of parallel power supply modules in the load change process.

Description

Technical Field

[0001] This invention relates to the field of power electronics technology, and in particular to a method and system for controlling the current sharing of multiple power supply modules in parallel. Background Technology

[0002] With the development of power electronics technology, modular power supply systems are widely used in fields such as communication power supplies, industrial control power supplies, new energy power generation and energy storage conversion systems, and rail transit power supply systems. To improve power supply capacity, system reliability, and ease of maintenance, multiple power modules are typically connected in parallel to the same output bus, allowing multiple modules to jointly supply power to the load. Compared to a single high-power power supply, the parallel connection of multiple power modules offers advantages such as flexible capacity expansion, better redundancy, and easier modular maintenance.

[0003] When multiple power modules operate in parallel, uneven output current distribution can easily occur due to differences in factors such as output voltage setpoints, sampling errors, device parameters, control response speeds, line impedances, temperature rise, and aging levels among the modules. If some power modules bear a large output current for an extended period, it may lead to increased temperature rise, decreased efficiency, premature triggering of protection mechanisms in those modules, and even affect the stability and reliability of the entire parallel power supply system. Therefore, in a multi-module power supply parallel current sharing control system, achieving current sharing control among the power modules is a crucial issue for ensuring stable system operation.

[0004] Existing current sharing control methods typically detect the output current of each power module and adjust the output of each power module based on the current difference, average current, or current sharing bus signal. This type of method can improve steady-state current distribution to some extent, but under conditions of frequent load changes, large differences in module response, or long-term changes in operating conditions, it may still be difficult to accurately reflect the current distribution characteristics of each power module during different load changes, thus limiting the targeting and stability of current sharing regulation. Summary of the Invention

[0005] To address the aforementioned technical problems, this invention proposes a method and system for current sharing control of parallel multi-module power supplies. By analyzing the output current timing data of multiple parallel power supply modules, the current distribution characteristics of each parallel power supply module during historical load changes are determined and used for current sharing compensation control, thereby improving the current sharing control effect of the parallel multi-module power supply control system.

[0006] To achieve the above objectives, the first aspect of the present invention provides a method for controlling the current sharing of multiple power supply modules in parallel, comprising: Collect the output current timing data of multiple parallel power supply modules, and calculate the current sharing residual of each parallel power supply module at different sampling times based on multiple sets of output current timing data; Identify multiple historical load events in multiple sets of output current timing data, and construct the residual trajectory of each parallel power module in each historical load event based on multiple current sharing residuals and preset residual analysis windows; Based on multiple residual trajectories, construct the residual slot occupancy trajectory of each parallel power module for each historical load event, and divide the multiple historical load events into load increase events and load decrease events; The average residual trajectory for each sampling moment in each type of historical load event is constructed based on multiple residual slot occupancy trajectories, and the slot occupancy distribution data of each parallel power module at each sampling moment is extracted. The identity slot trajectory of each parallel power module is constructed based on multiple average residual trajectories and slot occupancy distribution data. The current sharing compensation trajectory of each parallel power module is constructed based on multiple identity slot trajectories and used for current sharing control of multiple parallel power modules.

[0007] Preferably, the construction of the residual trajectory and the residual slot occupancy trajectory includes: The total output current of multiple parallel power supply modules is calculated at each sampling time based on multiple sets of output current timing data. The actual current share of each parallel power supply module is determined based on the ratio of the individual output current of the parallel power supply module to the total output current at different sampling times. The difference between the actual current share and the target individual current share is used as the current sharing residual of the parallel power supply module at the corresponding sampling time. Taking the event start point of each historical load event as the time alignment benchmark, multiple current sharing residuals of each parallel power supply module are extracted according to the preset residual analysis window to obtain the residual trajectory of each parallel power supply module in each historical load event. For each historical load event, at each sampling time, the current sharing residuals of multiple parallel power modules are sorted according to their numerical values ​​to obtain multiple residual slots. The residual slot number occupied by each parallel power module at different sampling times is recorded to form the residual slot occupancy trajectory of each parallel power module in each historical load event.

[0008] Preferably, constructing the average residual trajectory for each sampling moment in each type of historical load event based on multiple residual slot occupancy trajectories, and extracting the slot occupancy distribution data for each parallel power module at each sampling moment includes: Based on multiple load increase events and load decrease events, construct load increase event sets and load decrease event sets. For any event set, at each sampling time in the preset residual analysis window, obtain the current sharing residual corresponding to each residual slot for multiple historical load events in the event set, and average the multiple current sharing residuals for each residual slot to obtain the average residual amount. Based on the multiple average residual amounts, construct the average residual trajectory for multiple residual slots at each sampling time. For any parallel power supply module, within each set of multiple historical load events, the number of times the parallel power supply module occupies each residual slot at each sampling moment in the preset residual analysis window is counted. The slot occupancy ratio of each residual slot is determined by combining the number of historical load events, so as to form the slot occupancy distribution data of the parallel power supply module with respect to multiple residual slots at each sampling moment.

[0009] Preferably, constructing the identity slot trajectory for each parallel power module based on multiple average residual trajectories and slot occupancy distribution data includes: For any parallel power module in each type of historical load event, the multiple slot occupancy ratios of the parallel power module at each sampling time are obtained based on the slot occupancy distribution data, and the deviation of each slot occupancy ratio of the parallel power module from the reference slot occupancy ratio is determined, where the reference slot occupancy ratio is the reciprocal of the number of parallel power modules. The deviation of each residual slot is multiplied by the corresponding average residual, and the multiple product results of the parallel power supply module at the same sampling time are summed to obtain the identity slot value of the parallel power supply module at each sampling time. Based on the multiple identity slot values, the identity slot trajectory of the parallel power supply module at multiple sampling times is constructed.

[0010] Preferably, constructing a current sharing compensation trajectory for each parallel power module based on multiple identity slot trajectories, and using it for current sharing control of multiple parallel power modules includes: Construct the initial current sharing compensation trajectory for each parallel power module for each type of historical load event based on the inverse vector of the identity slot trajectory. Zero-sum processing is performed on multiple initial current sharing compensation trajectories of the same type of historical load event to obtain the current sharing compensation trajectory corresponding to each parallel power supply module; When a current load event matching any type of historical load event is identified in real time, multiple current sharing compensation trajectories are applied to multiple parallel power supply modules, with the event start point of the current load event as the time reference, in order to adjust the output current share of the multiple parallel power supply modules.

[0011] Preferably, multiple initial current sharing compensation trajectories for the same type of historical load event are subjected to zero-sum processing to obtain the current sharing compensation trajectory corresponding to each parallel power module, including: For multiple initial current sharing compensation trajectories in the same type of historical load event, the average value of multiple initial current sharing compensation amounts at the same sampling time is calculated. The average value is then subtracted from the initial current sharing compensation amount of the parallel power supply module at each sampling time to obtain the target current sharing compensation amount of the parallel power supply module at each sampling time. The target current sharing compensation amounts at multiple sampling times are combined to construct the current sharing compensation trajectory of each parallel power supply module under different types of historical load events.

[0012] A second aspect of the present invention provides a multi-module power supply parallel current sharing control system for implementing the above-mentioned multi-module power supply parallel current sharing control method, comprising: The current sharing residual calculation module is used to collect the output current timing data of multiple parallel power supply modules and calculate the current sharing residual of each parallel power supply module at different sampling times based on multiple sets of output current timing data. The load event trajectory construction module is used to identify multiple historical load events in multiple sets of output current timing data, and construct the residual trajectory of each parallel power supply module in each historical load event based on multiple current sharing residuals and preset residual analysis windows. The residual slot trajectory construction module is used to construct the residual slot occupancy trajectory of each parallel power module for each historical load event based on multiple residual trajectories, and to divide multiple historical load events into load increase events and load decrease events. The identity slot trajectory generation module is used to construct the average residual trajectory of each sampling time in each type of historical load event based on multiple residual slot occupancy trajectories, and to extract the slot occupancy distribution data of each parallel power module at each sampling time. The current sharing compensation control module is used to construct the identity slot trajectory of each parallel power module based on multiple average residual trajectories and slot occupancy distribution data, construct the current sharing compensation trajectory of each parallel power module based on multiple identity slot trajectories, and perform current sharing control on multiple parallel power modules.

[0013] The present invention has the following beneficial effects: This invention performs event-based processing on the output current timing data of multiple parallel power supply modules, constructs residual slot occupancy trajectories based on current sharing residuals, and further generates identity slot trajectories by combining the average residual trajectories in each type of historical load event and the slot occupancy distribution data of each parallel power supply module. This can reflect the stable deviation characteristics of each parallel power supply module from the target current distribution state during load changes. The current sharing compensation trajectory is constructed based on the identity slot trajectories and applied in the matched current load event, which can avoid the randomness caused by relying solely on instantaneous current difference for adjustment, and improve the targeting and stability of current sharing control of multiple parallel power supply modules during load changes. Attached Figure Description

[0014] Figure 1 This is a flowchart illustrating a method for controlling the parallel current sharing of multiple power supplies in an embodiment of the present invention.

[0015] Figure 2 This is a schematic diagram of a multi-module power supply parallel current sharing control system according to an embodiment of the present invention. Detailed Implementation

[0016] To enable those skilled in the art to better understand the technical solutions of this invention, the technical solutions of the embodiments of this invention will be clearly and completely described below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are only for explaining the invention and are not intended to limit the invention.

[0017] Please see Figure 1 The first aspect of this invention provides a method for controlling the current sharing of multiple power supply modules in parallel, the method comprising the following steps: Step S1: Collect the output current timing data of multiple parallel power supply modules, and calculate the current sharing residual of each parallel power supply module at different sampling times based on multiple sets of output current timing data.

[0018] Specifically, multiple parallel power modules are connected to the same output bus and supply power to the same load or the same load bus. This multi-parallel power module structure can be applied to scenarios requiring modular parallel power supply, such as communication power supplies, industrial equipment power supplies, new energy power generation conversion systems, energy storage conversion systems, and vehicle or rail transit auxiliary power supply systems. The parallel power modules can be of the same specifications or can be power modules with different rated output capabilities that participate in parallel power supply.

[0019] In one implementation, the output current of each parallel power supply module can be acquired by a current detection unit located on the output side of each parallel power supply module. The current detection unit can be a Hall current sensor, a sampling resistor, a current transformer, or an existing current detection circuit within the power supply module. To facilitate subsequent current distribution analysis at the same sampling time, the acquisition times of the output current timing data of each parallel power supply module can be aligned to acquire the output current timing data of each parallel power supply module.

[0020] After obtaining multiple sets of output current timing data, the individual output current of each parallel power supply module is acquired at the same sampling time. The individual output currents of multiple parallel power supply modules at the same sampling time are then summed to obtain the total output current of the multiple parallel power supply modules at that sampling time. Based on the ratio of the individual output current of each parallel power supply module at different sampling times to the total output current at the corresponding sampling time, the actual current share of each parallel power supply module is determined. The actual current share represents the proportion of the output current borne by the corresponding parallel power supply module in the total output current at the corresponding sampling time.

[0021] Based on this, the actual current share of each parallel power module is compared with the corresponding target individual current share, and the difference between the two is used as the current sharing residual of the parallel power module at the corresponding sampling time. The aforementioned target individual current share is used to represent the output current share that the parallel power module should bear under ideal current distribution conditions. When multiple parallel power modules are of the same specification, the target individual current share can be the reciprocal of the number of parallel power modules. When the rated output capabilities of multiple parallel power modules are different, the target individual current share can be determined by calculating the proportion of the rated output capability of the parallel power module in the sum of the rated output capabilities of the multiple parallel power modules.

[0022] The aforementioned current sharing residual does not represent the difference in the absolute value of the original output current between the parallel power supply modules, but rather the degree of deviation of the actual current share of the corresponding parallel power supply module from its target individual current share. When the current sharing residual is positive, it indicates that the output current share borne by the corresponding parallel power supply module at that sampling time is higher than the target individual current share; when the current sharing residual is negative, it indicates that the output current share borne by the corresponding parallel power supply module at that sampling time is lower than the target individual current share; when the current sharing residual is close to zero, it indicates that the output current distribution of the corresponding parallel power supply module at that sampling time is close to the ideal current sharing distribution state.

[0023] Step S2: Identify multiple historical load events in multiple sets of output current timing data, and construct the residual trajectory of each parallel power supply module in each historical load event based on multiple current sharing residuals and preset residual analysis windows.

[0024] Specifically, historical load events represent the load change process that occurred during the historical operation of a system composed of multiple parallel power supply modules. By identifying historical load events in the output current timing data and constructing residual trajectories based on these historical load events, the originally continuous operating data can be transformed into event-based data associated with the load change process.

[0025] In one implementation, for the identification of historical load events, the total output current can be calculated based on the individual output currents of multiple parallel power modules at the same sampling time, yielding total output current timing data. Change detection is then performed on this total output current timing data. When the change in total output current between consecutive sampling times, or the cumulative change in total output current across multiple sampling times within a preset detection interval, reaches a preset load change threshold, it is determined that a historical load event exists near the corresponding sampling time. This preset load change threshold can be pre-set based on the rated output current, sampling period, and allowable load fluctuation range of the multi-module power supply parallel current sharing control system.

[0026] When identifying historical load events, the sampling moment when the total output current begins to change continuously is determined as the event starting point of the historical load event. To avoid misidentifying transient noise or short-term sampling jitter as historical load events, multiple sampling points within a preset detection range of the total output current change can be further constrained to maintain the same direction of change, thereby obtaining a more accurate event starting point.

[0027] After identifying multiple historical load events, the current sharing residuals of each parallel power supply module are extracted according to a preset residual analysis window, using the event start point of each historical load event as the time alignment reference. This yields the residual trajectory of each parallel power supply module in each historical load event. The preset residual analysis window includes multiple sampling times aligned with the event start point, and includes a pre-event sampling interval before the event start point and a post-event sampling interval after the event start point, used to cover the change process of the current sharing residuals of each parallel power supply module before and after the occurrence of the historical load event.

[0028] Specifically, for any historical load event, the event start point is taken as the relative time zero point. A predetermined number of sampling points before and after the event start point are included in the same residual analysis window, and the current sharing residual of the corresponding sampling points is extracted from the current sharing residual time series data. For any parallel power supply module, the extracted current sharing residuals are arranged according to the chronological order of the relative sampling times within the residual analysis window, thus forming the residual trajectory of the parallel power supply module in that historical load event.

[0029] Step S3: Construct the residual slot occupancy trajectory of each parallel power module for each historical load event based on multiple residual trajectories, and divide the multiple historical load events into load increase events and load decrease events.

[0030] Specifically, this embodiment does not directly use the residual value of a single parallel power module as the basis for current sharing control. Instead, it constructs a residual slot occupancy trajectory based on the relative residual positions of each parallel power module in the same historical load event. This trajectory characterizes the change in the residual ranking position of the corresponding parallel power module before and after the historical load event. By dividing historical load events into load increase events and load decrease events, the current distribution characteristics under different load change directions can be obtained separately, avoiding the mixed processing of module performance during different types of load changes.

[0031] In one implementation, for the construction process of residual slot occupancy trajectory, for any historical load event, the starting point of the historical load event is used as the time alignment reference, and the current sharing residuals corresponding to multiple parallel power modules at each sampling time are selected sequentially in the preset residual analysis window.

[0032] For any sampling time within the same historical load event, the current sharing residuals of multiple parallel power supply modules at that sampling time are sorted according to their numerical values, resulting in multiple residual slots. These residual slots represent the deviation ranking of the multiple parallel power supply modules relative to their respective target individual current shares at that sampling time. If a parallel power supply module occupies a higher residual slot, it indicates that the parallel power supply module has a higher positive deviation from its target individual current share at that sampling time; if it occupies a lower residual slot, it indicates that the parallel power supply module has a lower or negative deviation from its target individual current share at that sampling time.

[0033] In one specific implementation, the position corresponding to the parallel power supply module with the largest current sharing residual at the same sampling time can be determined as the first residual slot, the position corresponding to the parallel power supply module with the second largest current sharing residual can be determined as the second residual slot, and so on, until the position corresponding to the parallel power supply module with the smallest current sharing residual is determined as the Nth residual slot, where N is the number of parallel power supply modules participating in parallel power supply, and the residual slot number occupied by each parallel power supply module at each sampling time is recorded.

[0034] For multiple sampling times within a preset residual analysis window, the residual slot numbers occupied by the same parallel power supply module at different sampling times form the residual slot occupancy trajectory of that parallel power supply module in that historical load event. Each historical load event can generate multiple residual slot occupancy trajectories, and each residual slot occupancy trajectory corresponds to one parallel power supply module. Compared to simply recording the magnitude of the current sharing residual, the residual slot occupancy trajectory further reflects the relative residual position changes of the parallel power supply module during the load event.

[0035] In one implementation, the distinction between load increase events and load decrease events is made by determining the event start point after determining the event type based on the direction of change of the total output current within a preset detection interval after the event start point. When the total output current increases relative to the current level before the event start point, the historical load event is determined as a load increase event; when the total output current decreases relative to the current level before the event start point, the historical load event is determined as a load decrease event.

[0036] Step S4: Construct the average residual trajectory for each sampling moment in each type of historical load event based on multiple residual slot occupancy trajectories, and extract the slot occupancy distribution data for each parallel power module at each sampling moment.

[0037] Specifically, after obtaining the residual slot occupancy trajectory, this embodiment further constructs multiple average residual trajectories for each type of historical load event and extracts multiple sets of slot occupancy distribution data for each parallel power supply module. The average residual trajectory is used to characterize the residual level corresponding to different residual slots in the corresponding category of historical load events, and the slot occupancy distribution data is used to characterize the distribution of each parallel power supply module occupying different residual slots at different sampling times.

[0038] In one implementation, the process of constructing the average residual trajectory and extracting slot occupancy distribution data specifically includes: Based on the event type of historical load events, multiple historical load events are divided into a load increase event set and a load decrease event set. The load increase event set includes multiple load increase events, and the load decrease event set includes multiple load decrease events.

[0039] For any set of events, at each sampling time within the preset residual analysis window, the current sharing residuals corresponding to multiple historical load events in the event set at each residual slot are obtained. Specifically, for multiple historical load events in the same event set, at the same sampling time, the first residual slot corresponds to the slot with the highest current sharing residual at that sampling time, the second residual slot corresponds to the slot with the second highest current sharing residual at that sampling time, and so on. The current sharing residuals of multiple historical load events in the same event set at different residual slots are obtained at each sampling time, and the multiple current sharing residuals corresponding to each residual slot are averaged to obtain the average residual amount corresponding to each residual slot at each sampling time. According to the arrangement order of the residual slots, the average residual amounts corresponding to multiple residual slots at the same sampling time are arranged to obtain the average residual trajectory of multiple residual slots at each sampling time.

[0040] Furthermore, for any parallel power supply module, the slot occupancy status of the parallel power supply module is statistically analyzed in both the load increase event set and the load decrease event set. Specifically, for any event set, at each sampling time within the preset residual analysis window, the number of times the parallel power supply module occupies each residual slot in multiple historical load events of that event set is counted.

[0041] After obtaining the number of times each residual slot is occupied, the slot occupancy ratio is determined based on the number of occupancy counts and the number of historical load events in the event set. That is, the ratio between the number of times the parallel power module occupies the residual slot in the event set at the sampling time and the total number of historical load events in the event set.

[0042] For the same parallel power supply module and the same sampling time, the slot occupancy ratio of the parallel power supply module in multiple residual slots is calculated. The slot occupancy ratios of multiple slots together constitute the slot occupancy distribution data of the parallel power supply module with respect to multiple residual slots at each sampling time. This data can reflect the statistical distribution of the residual position of each parallel power supply module relative to other parallel power supply modules during a specific type of load change.

[0043] Using the above method, two types of data can be obtained for the load increase event set and the load decrease event set respectively. The first is the average residual trajectory of multiple residual slots at each sampling time under the corresponding event set, and the second is the slot occupancy distribution data of each parallel power module at each sampling time under the corresponding event set.

[0044] Step S5: Construct the identity slot trajectory of each parallel power module based on multiple average residual trajectories and slot occupancy distribution data, construct the current sharing compensation trajectory of each parallel power module based on multiple identity slot trajectories, and use it to perform current sharing control on multiple parallel power modules.

[0045] Specifically, by combining the average residual trajectory with slot occupancy distribution data, the identity slot trajectory of each parallel power supply module in historical load events can be obtained to characterize the correlation between each parallel power supply module and the residual slot during load changes. Based on multiple identity slot trajectories, current sharing compensation trajectories corresponding to different types of historical load events for the parallel power supply modules are further constructed. During subsequent operation, current sharing control of multiple parallel power supply modules is performed based on these compensation trajectories. This allows the current sharing control process to not only consider the instantaneous output current differences of each parallel power supply module but also to incorporate the slot occupancy characteristics of each parallel power supply module during historical load changes, thereby adjusting the output current share of multiple parallel power supply modules. This improves the current sharing control effect of the multi-module parallel power supply current sharing control system under different load change conditions.

[0046] In one implementation, constructing the identity slot trajectory for each parallel power module specifically includes: For any parallel power supply module in each type of historical load event, based on the slot occupancy distribution data corresponding to the parallel power supply module, obtain the slot occupancy ratio of the parallel power supply module with respect to multiple residual slots at each sampling time in the preset residual analysis window.

[0047] In this embodiment, if the number of parallel power supply modules participating in parallel power supply is N, then each sampling time includes N residual slots. The reciprocal of the number of parallel power supply modules can be determined as the reference slot occupancy ratio, that is, the reference slot occupancy ratio is (1 / N). The reference slot occupancy ratio is used to represent the theoretical balanced ratio of a parallel power supply module occupying any residual slot in multiple historical load events when there is no stable slot occupancy bias.

[0048] For any parallel power supply module, any sampling time, and any residual slot, the slot occupancy ratio of the parallel power supply module in that residual slot is compared with the reference slot occupancy ratio, and the difference between the two is determined as the deviation of the parallel power supply module from the residual slot at the sampling time. If the deviation is positive, it means that the frequency of the parallel power supply module occupying the residual slot at that sampling time is higher than the reference slot occupancy ratio; if the deviation is negative, it means that the frequency of the parallel power supply module occupying the residual slot at that sampling time is lower than the reference slot occupancy ratio.

[0049] Further, the average residual trajectory for each sampling time under the corresponding event category is obtained. For each residual slot at each sampling time, the deviation of the parallel power supply module from that residual slot is multiplied by the average residual amount corresponding to that residual slot to obtain the product result corresponding to that residual slot. After obtaining the product results corresponding to multiple residual slots at the same sampling time, the multiple product results at the same sampling time are summed to obtain the identity slot value of the parallel power supply module at that sampling time. This value represents the comprehensive residual performance of the parallel power supply module under the corresponding event category and the corresponding sampling time, formed by the deviation of its slot occupancy distribution from the baseline slot occupancy ratio.

[0050] The slot identity value in this application is not directly determined by the single-cycle current sharing residual of the parallel power module, but is determined by two factors: first, the distribution deviation of the parallel power module occupying each residual slot in multiple historical load events; and second, the average residual level exhibited by each residual slot at the relative sampling time of the load event. Therefore, the slot identity value can characterize whether the parallel power module stably tends to occupy high-residual-slot or low-residual-slot in similar historical load events.

[0051] Using the above method, the identity slot value of each parallel power module at multiple sampling times can be calculated sequentially, and the multiple identity slot values ​​of each parallel power module can be arranged in chronological order to obtain the identity slot trajectory of each parallel power module under the corresponding event category.

[0052] In one specific implementation, when constructing the current sharing compensation trajectory of the parallel power supply modules, the inverse vector of the identity slot trajectory can be extracted first to obtain the initial current sharing compensation trajectory of each parallel power supply module for each type of historical load event.

[0053] In a more preferred embodiment, to avoid excessive compensation due to historical data fluctuations or excessively large identity slot values, and to improve the stability of the current sharing control process, after extracting the inverse vector of the identity slot trajectory, the inverse vector of the identity slot trajectory is scaled according to a preset compensation ratio coefficient to limit the compensation amount. The limited inverse vector is used as the initial current sharing compensation trajectory of the parallel power supply module, where each element represents the initial current sharing compensation amount at different sampling times.

[0054] To avoid an overall increase or decrease in the current sharing compensation of multiple parallel power modules, a zero-sum processing method is applied to multiple initial current sharing compensation values ​​under the same event category and the same sampling time. Specifically, the average value of the initial current sharing compensation values ​​of multiple parallel power modules at each sampling time is calculated, and the calculated average value is subtracted from the initial current sharing compensation value of the parallel power modules at each sampling time to obtain the target current sharing compensation value of the parallel power modules at each sampling time. The target current sharing compensation values ​​of each parallel power module at multiple sampling times are combined to construct the current sharing compensation trajectory of each parallel power module under different types of historical load events.

[0055] During the real-time operation of the multi-module power supply parallel current sharing control system, real-time output current timing data of multiple parallel power supply modules are collected, and the current load event is identified based on the change in total output current. When the real-time identified current load event matches any of the event categories of load increase or load decrease, multiple current sharing compensation trajectories corresponding to that event category are invoked.

[0056] When applying the current sharing compensation trajectory, the event start point of the current load event is used as the time reference, and the event start point of the current load event is determined as the relative sampling position zero point. The target current sharing compensation amount of each parallel power supply module is read from the current sharing compensation trajectory of the corresponding event category, and the read current sharing compensation amount is applied to the corresponding parallel power supply module to adjust the output current share of multiple parallel power supply modules.

[0057] In this way, when a current load event occurs, each parallel power supply module does not compensate solely based on the instantaneous current difference, but rather adjusts according to the current sharing compensation trajectory generated from the identity slot trajectory formed in historical similar load events. This allows multiple parallel power supply modules to perform targeted current sharing corrections based on historical current distribution characteristics during load changes, thereby reducing the likelihood of some parallel power supply modules occupying high-residual-error slots or low-residual-error slots for extended periods, and improving the current sharing control effect of the multi-module parallel power supply current sharing control system during load changes.

[0058] Please see Figure 2 The second aspect of this invention provides a multi-module power supply parallel current sharing control system, specifically used to implement the above-mentioned multi-module power supply parallel current sharing control method. The system includes: The current sharing residual calculation module is used to collect the output current timing data of multiple parallel power supply modules and calculate the current sharing residual of each parallel power supply module at different sampling times based on multiple sets of output current timing data. The load event trajectory construction module is used to identify multiple historical load events in multiple sets of output current timing data, and construct the residual trajectory of each parallel power supply module in each historical load event based on multiple current sharing residuals and preset residual analysis windows. The residual slot trajectory construction module is used to construct the residual slot occupancy trajectory of each parallel power module for each historical load event based on multiple residual trajectories, and to divide multiple historical load events into load increase events and load decrease events. The identity slot trajectory generation module is used to construct the average residual trajectory of each sampling time in each type of historical load event based on multiple residual slot occupancy trajectories, and to extract the slot occupancy distribution data of each parallel power module at each sampling time. The current sharing compensation control module is used to construct the identity slot trajectory of each parallel power module based on multiple average residual trajectories and slot occupancy distribution data, construct the current sharing compensation trajectory of each parallel power module based on multiple identity slot trajectories, and perform current sharing control on multiple parallel power modules.

[0059] The above are merely specific embodiments of the present invention, enabling those skilled in the art to understand or implement the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art. Parts not described in detail in this specification are prior art known to those skilled in the art.

Claims

1. A method for controlling current sharing in parallel connection of multi-module power supplies, characterized in that, include: Collect the output current timing data of multiple parallel power supply modules, and calculate the current sharing residual of each parallel power supply module at different sampling times based on multiple sets of output current timing data; Identify multiple historical load events in multiple sets of output current timing data, and construct the residual trajectory of each parallel power module in each historical load event based on multiple current sharing residuals and preset residual analysis windows; Based on multiple residual trajectories, construct the residual slot occupancy trajectory of each parallel power module for each historical load event, and divide the multiple historical load events into load increase events and load decrease events; The average residual trajectory for each sampling moment in each type of historical load event is constructed based on multiple residual slot occupancy trajectories, and the slot occupancy distribution data of each parallel power module at each sampling moment is extracted. The identity slot trajectory of each parallel power module is constructed based on multiple average residual trajectories and slot occupancy distribution data. The current sharing compensation trajectory of each parallel power module is constructed based on multiple identity slot trajectories and used for current sharing control of multiple parallel power modules.

2. The multi-module power supply parallel current sharing control method according to claim 1, characterized in that, The construction of residual trajectories and residual slot occupancy trajectories includes: The total output current of multiple parallel power supply modules is calculated at each sampling time based on multiple sets of output current timing data. The actual current share of each parallel power supply module is determined based on the ratio of the individual output current of the parallel power supply module to the total output current at different sampling times. The difference between the actual current share and the target individual current share is used as the current sharing residual of the parallel power supply module at the corresponding sampling time. Taking the event start point of each historical load event as the time alignment benchmark, multiple current sharing residuals of each parallel power supply module are extracted according to the preset residual analysis window to obtain the residual trajectory of each parallel power supply module in each historical load event. For each historical load event, at each sampling time, the current sharing residuals of multiple parallel power modules are sorted according to their numerical values ​​to obtain multiple residual slots. The residual slot number occupied by each parallel power module at different sampling times is recorded to form the residual slot occupancy trajectory of each parallel power module in each historical load event.

3. The multi-module power supply parallel current sharing control method according to claim 2, characterized in that, Based on multiple residual slot occupancy trajectories, the average residual trajectory for each sampling time in each type of historical load event is constructed, and the slot occupancy distribution data for each parallel power module at each sampling time is extracted, including: Based on multiple load increase events and load decrease events, construct load increase event sets and load decrease event sets. For any event set, at each sampling time in the preset residual analysis window, obtain the current sharing residual corresponding to each residual slot for multiple historical load events in the event set, and average the multiple current sharing residuals for each residual slot to obtain the average residual amount. Based on the multiple average residual amounts, construct the average residual trajectory for multiple residual slots at each sampling time. For any parallel power supply module, within each set of multiple historical load events, the number of times the parallel power supply module occupies each residual slot at each sampling moment in the preset residual analysis window is counted. The slot occupancy ratio of each residual slot is determined by combining the number of historical load events, so as to form the slot occupancy distribution data of the parallel power supply module with respect to multiple residual slots at each sampling moment.

4. The multi-module power supply parallel current sharing control method according to claim 3, characterized in that, The identity slot trajectory of each parallel power module is constructed based on multiple average residual trajectories and slot occupancy distribution data, including: For any parallel power module in each type of historical load event, the multiple slot occupancy ratios of the parallel power module at each sampling time are obtained based on the slot occupancy distribution data, and the deviation of each slot occupancy ratio of the parallel power module from the reference slot occupancy ratio is determined, where the reference slot occupancy ratio is the reciprocal of the number of parallel power modules. The deviation of each residual slot is multiplied by the corresponding average residual, and the multiple product results of the parallel power supply module at the same sampling time are summed to obtain the identity slot value of the parallel power supply module at each sampling time. Based on the multiple identity slot values, the identity slot trajectory of the parallel power supply module at multiple sampling times is constructed.

5. The multi-module power supply parallel current sharing control method according to claim 4, characterized in that, The current sharing compensation trajectory for each parallel power module is constructed based on multiple identity slot trajectories, and is used for current sharing control of multiple parallel power modules, including: Construct the initial current sharing compensation trajectory for each parallel power module for each type of historical load event based on the inverse vector of the identity slot trajectory. Zero-sum processing is performed on multiple initial current sharing compensation trajectories of the same type of historical load event to obtain the current sharing compensation trajectory corresponding to each parallel power supply module; When a current load event matching any type of historical load event is identified in real time, multiple current sharing compensation trajectories are applied to multiple parallel power supply modules, with the event start point of the current load event as the time reference, in order to adjust the output current share of the multiple parallel power supply modules.

6. The multi-module power supply parallel current sharing control method according to claim 5, characterized in that, Zero-sum processing is performed on multiple initial current sharing compensation trajectories for the same type of historical load event to obtain the current sharing compensation trajectory corresponding to each parallel power module, including: For multiple initial current sharing compensation trajectories in the same type of historical load event, the average value of multiple initial current sharing compensation amounts at the same sampling time is calculated. The average value is then subtracted from the initial current sharing compensation amount of the parallel power supply module at each sampling time to obtain the target current sharing compensation amount of the parallel power supply module at each sampling time. The target current sharing compensation amounts at multiple sampling times are combined to construct the current sharing compensation trajectory of each parallel power supply module under different types of historical load events.

7. A multi-module power supply parallel current sharing control system, characterized in that, The system is used to implement the multi-module power supply parallel current sharing control method according to any one of claims 1-6, including: The current sharing residual calculation module is used to collect the output current timing data of multiple parallel power supply modules and calculate the current sharing residual of each parallel power supply module at different sampling times based on multiple sets of output current timing data. The load event trajectory construction module is used to identify multiple historical load events in multiple sets of output current timing data, and construct the residual trajectory of each parallel power supply module in each historical load event based on multiple current sharing residuals and preset residual analysis windows. The residual slot trajectory construction module is used to construct the residual slot occupancy trajectory of each parallel power module for each historical load event based on multiple residual trajectories, and to divide multiple historical load events into load increase events and load decrease events. The identity slot trajectory generation module is used to construct the average residual trajectory of each sampling time in each type of historical load event based on multiple residual slot occupancy trajectories, and to extract the slot occupancy distribution data of each parallel power module at each sampling time. The current sharing compensation control module is used to construct the identity slot trajectory of each parallel power module based on multiple average residual trajectories and slot occupancy distribution data, construct the current sharing compensation trajectory of each parallel power module based on multiple identity slot trajectories, and perform current sharing control on multiple parallel power modules.

Images