A communication base station power load balancing data processing method and system

Abstract

This invention discloses a data processing method and system for power load balancing in communication base stations, relating to the field of data processing technology. The method includes: obtaining transient current step values ​​based on data scheduling parameters and energy consumption benchmarks, and determining first and second response time-domain windows in conjunction with the actual output current; obtaining the minimum voltage drop point based on the bus electrical state sequence, thereby dividing the two windows into pre- and post-response stages; calculating the difference between the benchmark and the actual voltage drop slope in the pre-response stage; if the difference exceeds a tolerance threshold, adjusting the virtual output impedance parameters based on this difference and the step value; simultaneously, obtaining the thermal resistance characteristics and conduction loss parameters of the load distribution unit to calculate the current thermal tolerance margin, and allocating the total load current accordingly to generate the target output current. This invention has the advantages of fast transient response, reliable stage division, and balanced thermal distribution.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, specifically to a data processing method and system for power load balancing of communication base stations. Background Technology

[0002] With the nonlinear bursts of traffic to communication base stations (such as large bandwidth requests or high-frequency resource scheduling), the transient power demand of radio frequency units will change abruptly. The existing base station power control architecture has the risk of excessive dynamic bus voltage drop and steady-state thermal runaway due to the lack of decoupling between passive feedback response and physical mechanism.

[0003] Specifically, existing closed-loop control relies on voltage or current sampling feedback from the power supply side, making it difficult to detect sudden current jumps caused by baseband data scheduling in advance. This sampling and loop delay causes the bus filter capacitor to be heavily consumed before the inductor current rises, easily triggering the base station's low-voltage reset protection. Furthermore, in handling the response process, existing solutions typically fail to effectively distinguish between the transient voltage drop phase and the steady-state current sharing phase, only performing global scaling in the time domain. This ignores the physical boundary transition from bus capacitor discharge support to inductor current takeover, leading to a mismatch between the regulation strategy and the electrical characteristics of the power supply hardware. Moreover, after entering the steady-state distribution phase, existing load balancing methods typically distribute loads based on the average output current of each parallel module or a fixed capacity ratio, failing to fully consider the changes in thermal resistance and conduction losses caused by differences in heat dissipation conditions, filter blockage, or device aging in actual operation of each sub-module. When a base station is under heavy load, this fixed current sharing strategy will cause the power module with deteriorating heat dissipation to continue to bear the same high current load. Its internal junction temperature will continue to rise, increasing the thermal stress loss of power devices, increasing the risk of hardware damage and communication interruption, and causing the communication base station power supply system to face the problem of not being able to balance transient impacts and long-term heat distribution. Summary of the Invention

[0004] In view of the technical problems described in the background art, the present invention provides a method and system for processing power load balancing data of communication base stations.

[0005] A data processing method for power load balancing in a communication base station includes: acquiring data scheduling parameters and energy consumption benchmarks, and acquiring transient current step values ​​based on the data scheduling parameters and energy consumption benchmarks; determining the starting point of a first response time-domain window based on the transient current step values, acquiring the actual output current, and determining a second response time-domain window based on the actual output current; acquiring a bus electrical state sequence, and acquiring the minimum point of bus voltage drop based on the bus electrical state sequence; dividing the first response time-domain window into a first pre-response stage and a first post-response stage based on the minimum point of bus voltage drop, and dividing the second response time-domain window into a second pre-response stage. The system consists of a first pre-response stage and a second post-response stage. The reference voltage drop slope of the first pre-response stage and the actual voltage drop slope of the second pre-response stage are obtained respectively. The difference between the reference voltage drop slope and the actual voltage drop slope is then calculated. If the difference in the pre-response slope exceeds a tolerance threshold, the virtual output impedance parameters of the second pre-response stage are adjusted based on the difference in the pre-response slope and the transient current step. The thermal resistance characteristic parameters and conduction loss parameters of the load distribution unit are obtained. The current thermal tolerance margin is then calculated based on the thermal resistance characteristic parameters and conduction loss parameters. The total load current is distributed based on the current thermal tolerance margin to generate the target output current.

[0006] Optionally, data scheduling parameters and energy consumption benchmarks are obtained, and transient current step is obtained based on the data scheduling parameters and energy consumption benchmarks, including: obtaining burst data increments as data scheduling parameters and obtaining single-bit dynamic energy consumption as energy consumption benchmarks; obtaining the current bus voltage and scheduling duration; and obtaining transient current step based on single-bit dynamic energy consumption, burst data increments, current bus voltage and scheduling duration.

[0007] Optionally, determining the starting point of the first response time-domain window based on the transient current step, obtaining the actual output current, and determining the second response time-domain window based on the actual output current includes: taking the expected moment of generating the transient current step as the starting point of the first response time-domain window; obtaining the current trajectory interval where the actual current change occurs based on the actual output current; and establishing the current trajectory interval as the second response time-domain window.

[0008] Optionally, the bus electrical state sequence is obtained, and the minimum point of bus voltage drop is obtained based on the bus electrical state sequence, including: obtaining the first and second derivatives of the bus electrical state sequence as a function of time; obtaining the physical moment when the first derivative is zero and the second derivative is greater than zero; and calibrating the physical moment as the minimum point of bus voltage drop indicating the end of capacitor discharge support.

[0009] Optionally, based on the minimum point of the bus voltage drop, the first response time domain window is divided into a first pre-response stage and a first post-response stage, and the second response time domain window is divided into a second pre-response stage and a second post-response stage, including: defining the time interval from the initial jump point to the minimum point of the bus voltage drop as the first pre-response stage and the second pre-response stage; and defining the time interval from the minimum point of the bus voltage drop to the moment when the current output tends to stabilize as the first post-response stage and the second post-response stage.

[0010] Optionally, the virtual output impedance parameters of the second pre-response stage are adjusted based on the difference in the pre-response slope and the transient current step, including: calculating the impedance compensation coefficient based on the transient current step and the difference in the pre-response slope; and reducing the virtual output impedance parameters in the control loops of each load distribution unit in the power supply state based on the impedance compensation coefficient, so as to globally improve the rate of change of transient inductor current over time.

[0011] Optionally, the thermal resistance characteristic parameters and conduction loss parameters of the load distribution unit are obtained, and a current thermal capacity margin is obtained based on the thermal resistance characteristic parameters and conduction loss parameters. The total load current is then distributed according to the current thermal capacity margin to generate a target output current, including: obtaining the current junction temperature and a safe junction temperature threshold; obtaining a temperature margin based on the safe junction temperature threshold and the current junction temperature; obtaining the current thermal capacity margin of each load distribution unit based on the temperature margin, thermal resistance characteristic parameters, and conduction loss parameters; performing a square root operation on the current thermal capacity margin of each load distribution unit to obtain the corresponding linear current margin; obtaining the distribution weight based on the ratio of the linear current margin of each load distribution unit to the sum of the linear current margins of all load distribution units; and weighted distribution of the total load current according to the distribution weight to obtain the target output current of the corresponding load distribution unit.

[0012] A data processing system for power load balancing of a communication base station is also provided, comprising: a parameter acquisition and window establishment module, used to acquire data scheduling parameters and energy consumption benchmarks, and acquire transient current step values ​​based on the data scheduling parameters and energy consumption benchmarks; determine the starting jump point of the first response time domain window based on the transient current step value, acquire the actual output current, and determine the second response time domain window based on the actual output current; a physical segmentation module, used to acquire the bus electrical state sequence, and acquire the minimum value point of the bus voltage drop based on the bus electrical state sequence; divide the first response time domain window into a first pre-response stage and a first post-response stage based on the minimum value point of the bus voltage drop, and divide the second response time domain window into a second... The system comprises a pre-response stage and a second post-response stage; a deviation evaluation module, used to obtain the reference voltage drop slope of the first pre-response stage and the actual voltage drop slope of the second pre-response stage, and to obtain the difference in pre-response slope based on the reference voltage drop slope and the actual voltage drop slope; and a step reduction adjustment module, used to adjust the virtual output impedance parameter of the second pre-response stage based on the difference in pre-response slope and the transient current step if the difference in pre-response slope exceeds the tolerance threshold, to obtain the thermal resistance characteristic parameter and conduction loss parameter of the load distribution unit, and to obtain the current thermal capacity margin based on the thermal resistance characteristic parameter and conduction loss parameter, and to distribute the total load current based on the current thermal capacity margin to generate the target output current.

[0013] Optionally, the parameter acquisition and window establishment module is also used to: acquire burst data increments as data scheduling parameters and acquire single-bit dynamic energy consumption as energy consumption benchmarks; acquire the current bus voltage and scheduling duration; and acquire transient current step based on single-bit dynamic energy consumption, burst data increments, current bus voltage, and scheduling duration.

[0014] Optionally, the order reduction adjustment module is also used to: obtain the current junction temperature and the safe junction temperature threshold; obtain the temperature margin based on the safe junction temperature threshold and the current junction temperature; obtain the current thermal tolerance margin of each load distribution unit based on the temperature margin, thermal resistance characteristic parameters, and conduction loss parameters; perform a square root operation on the current thermal tolerance margin of each load distribution unit to obtain the corresponding linear current margin; obtain the allocation weight based on the ratio of the linear current margin of each load distribution unit to the sum of the linear current margins of all load distribution units; and perform weighted allocation of the total load current based on the allocation weight to obtain the target output current of the corresponding load distribution unit.

[0015] The beneficial effects of this invention are reflected in: In the overall power load balancing data processing method for communication base stations, firstly, a feedforward model is constructed by introducing baseband-side data scheduling parameters and energy consumption benchmarks. This changes the existing feedback method where the power supply waits for the bus voltage drop before responding, enabling the system to quantify transient current steps and determine the initial jump point in advance, thus compensating for the sampling delay in the feedback loop. Furthermore, the derivative characteristics of the bus electrical state sequence are used to determine the minimum point of the bus voltage drop. Using this physical anchor point, the response window is divided into a pre-response stage supported by capacitor discharge and a post-response stage supported by inductor current takeover, avoiding the misalignment problem caused by directly scaling on the time axis. This ensures that the stage segmentation boundary aligns with the electrical evolution process. Furthermore, it enables independent evaluation of the deviation in the voltage drop slope in the first stage and the thermal resistance operating parameters in the second stage, providing data for subsequent adjustments. Furthermore, in the pre-response stage, by dynamically reducing the virtual output impedance parameter based on the step value, the rate of change of the transient inductor current is improved, suppressing the voltage drop amplitude of the bus. In the post-response stage, a current thermal tolerance margin model is established based on the real-time junction temperature, thermal resistance, and conduction loss of each module, allowing modules with good heat dissipation and high temperature margins to share more load, achieving adaptive balance in heat distribution and reducing the risk of thermal runaway caused by heat accumulation in local modules. Attached Figure Description

[0016] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the accompanying drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. In all the drawings, similar elements or parts are generally identified by similar reference numerals. In the drawings, the elements or parts are not necessarily drawn to scale.

[0017] Figure 1 This is a schematic diagram illustrating the steps of the communication base station power load balancing data processing method of the present invention; Figure 2 This is a schematic diagram of a portion of steps S1 in the communication base station power load balancing data processing method of the present invention; Figure 3 This is a schematic diagram of a portion of step S2 in the communication base station power load balancing data processing method of the present invention; Figure 4 This is a schematic diagram of a portion of step S3 in the communication base station power load balancing data processing method of the present invention; Figure 5 This is a schematic diagram of a portion of step S4 in the communication base station power load balancing data processing method of the present invention. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0019] Therefore, the following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.

[0020] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, the terms "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0021] This invention provides a data processing method for power load balancing in communication base stations, such as... Figure 1 As shown, in one specific embodiment, the method includes: S1. Obtain data scheduling parameters and energy consumption benchmark, and obtain transient current step based on data scheduling parameters and energy consumption benchmark; determine the starting jump point of the first response time domain window based on transient current step, obtain the actual output current, and determine the second response time domain window based on actual output current. S2. Obtain the bus electrical state sequence and obtain the minimum value point of bus voltage drop based on the bus electrical state sequence; based on the minimum value point of bus voltage drop, divide the first response time domain window into the first pre-response stage and the first post-response stage, and divide the second response time domain window into the second pre-response stage and the second post-response stage. S3. Obtain the reference voltage drop slope of the first pre-response stage and the actual voltage drop slope of the second pre-response stage respectively, and obtain the difference in pre-response slope based on the reference voltage drop slope and the actual voltage drop slope. S4. If the difference in the slope of the preceding response exceeds the tolerance threshold, the virtual output impedance parameter of the second preceding response stage is adjusted according to the difference in the slope of the preceding response and the transient current step, the thermal resistance characteristic parameter and conduction loss parameter of the load distribution unit are obtained, and the current thermal tolerance margin is obtained according to the thermal resistance characteristic parameter and conduction loss parameter. The total load current is distributed according to the current thermal tolerance margin to generate the target output current.

[0022] In this embodiment, it should be noted that in S1, the feedback lag limitation of existing power supplies passively waiting for voltage drops is addressed by establishing a physical mapping between baseband data scheduling and power supply feedforward, thereby pre-locking the initial demand for transient current. Among these, the energy consumption reference... This statistical characteristic value is determined by retrieving the operation logs of the same type of communication base station equipment within a preset historical period, calculating the ratio of the dynamic total power consumption increment of the radio frequency unit under different service loads to the corresponding total amount of scheduled data, and then performing mean filtering. Specifically, the method involves continuously collecting data traffic and bus current fluctuation data over multiple baseband processing cycles, removing the static power consumption baseline, and then calculating the average energy consumed per bit. For example, retrieving operation data from multiple base stations over the past 48 hours, the measured cumulative dynamic total power consumption is 1200J, corresponding to a total amount of processed data of... bits, then the calculated energy consumption baseline is This parameter reflects the inherent energy efficiency level under a specific hardware architecture.

[0023] In specific application scenarios, when the base station processing unit anticipates a burst of 4×10^7 bits of data increment within the next 1ms wireless scheduling frame, it no longer relies on bus voltage drop sampling. Instead, it combines the single-bit dynamic energy consumption of 30nJ / bit and the current bus voltage of 48V to directly calculate the transient current step of approximately 25A generated by the burst data. This calculation logic directly quantifies the data scheduling behavior of communication into the physical energy demand of the power supply, and uses this expected moment as the starting point of the first response time-domain window. Through this cross-domain feedforward mechanism, the power supply can obtain accurate dynamic warnings before the physical load experiences a substantial impact, changing the response lag problem caused by sampling and loop delays in existing feedback control. The beneficial effect of this step is that it provides a time window and quantification indicators for the transient support of the power module in advance, enabling subsequent impedance adjustment and energy injection to be aligned with actual load changes in time. This reduces the risk of excessive bus voltage drops caused by untimely response and enhances the overall stability of the power supply architecture when dealing with bursts of high-bandwidth services.

[0024] In S2, the physical boundary of the dynamic response stage is established by the actual electrical evolution characteristics of the bus voltage, rather than a simple fixed time length. In the above scenario, when the bus voltage is impacted by a 25A transient step current and starts to drop from 48V, the discrete sequence of voltage is monitored in real time and its derivative changes are calculated. When the voltage drop lasts for about 200μs to 47.1V and begins to rise, the physical moment when the first derivative is 0 and the second derivative is greater than 0 is accurately captured, i.e., the minimum point of the bus voltage drop. With this point as the boundary, the entire response process is strictly divided into a pre-response stage, which mainly controls capacitor discharge and voltage drop amplitude, and a post-response stage, which mainly involves inductor current ramp-up and the establishment of a new steady-state distribution. This division logic is based on the physical process of energy transfer within the power supply, avoiding the timing misalignment caused by artificially set time windows or simple peak-and-trough segmentation in existing methods. Its beneficial effects are reflected in the realization of refined physical order reduction of complex dynamic processes, ensuring that the transient support strategy in the front end and the thermal steady-state current sharing strategy in the back end can each play their respective roles in their corresponding electrical stages without interfering with each other. This partitioning mechanism based on physical evolution boundaries improves the pertinence and accuracy of subsequent deviation assessment and parameter adjustment, making the control logic more consistent with the actual operating laws of power electronic hardware.

[0025] In S3, the goal is to independently quantify the transient voltage support deviation in the pre-response stage and the module operating parameters in the post-response stage within the physical phase segmented in S2, providing data support for subsequent closed-loop regulation. The tolerance threshold is determined based on the voltage sag slope distribution pattern of the base station power system under normal operating conditions. Specifically, it involves collecting a large number of historical transient response samples, calculating the probability density of the voltage sag slope for each sample, and using deviation points outside the 95% confidence interval as the boundary for triggering regulation. For example, analysis of historical operating data reveals that the voltage sag slope under normal fluctuations is mainly concentrated in... to In order to avoid frequent triggering of misadjustments and to ensure the system's sensitivity to deep drops, the tolerance threshold is set to [value missing]. If the actual slope difference exceeds this boundary, the current transient support strength of the system is determined to be insufficient.

[0026] In the pre-response phase, the reference voltage drop slope (e.g., -4V / ms) is compared with the actual monitored voltage drop slope (e.g., -4.5V / ms), and the 0.5V / ms difference in the pre-response slope is calculated. It is then determined whether this difference exceeds the preset 0.2V / ms tolerance threshold. This calculation logic objectively reflects the gap between the current power module's transient response speed and safety requirements. Simultaneously, in the post-response phase, the current junction temperature, ambient temperature, and instantaneous output current of the four parallel power modules are collected in real time. For example, the junction temperature of module 1 is 75℃, and that of module 3 is 95℃. This step aims to comprehensively understand the real-time electrical and thermodynamic states at different response stages, avoiding the drawbacks of relying solely on a single voltage or current threshold for coarse judgment in existing solutions. Its beneficial effect is providing multi-dimensional, high-precision state input for subsequent adjustment phases, enabling accurate identification of whether the problem lies in insufficient transient support in the pre-response phase or potential thermal distribution issues in the post-response phase. This triggers corresponding directional correction mechanisms, improving the scientific rigor and effectiveness of the overall control strategy.

[0027] In S4, based on the evaluation results of S3, directional regulation based on thermoelectric coupling is performed to achieve synergistic optimization of transient support and steady-state current sharing. For the 0.5V / ms slope deviation in the pre-response phase, based on the 25A transient current step, the virtual output impedance parameter in the power module voltage control loop with hysteresis is reduced from 10mΩ to 6mΩ, thereby increasing the rise rate of the inductor current and pulling the voltage minimum point under similar fluctuations to above 47.5V. In the steady-state distribution phase, simple average current sharing is no longer used. Instead, the current thermal capacity margin of each module is calculated based on physical parameters such as the highest safe junction temperature (e.g., 125℃), the current junction temperature (e.g., 75℃ for module 1, 95℃ for module 3), thermal resistance (e.g., 0.6℃ / W and 0.9℃ / W), and conduction loss parameters (in this embodiment, these are represented as the equivalent on-resistance of the power switching devices, e.g., 4mΩ). Calculations show that the margin of module 1 (approximately 20833A²) is significantly higher than that of module 3 (approximately 8333A²). Based on this, a weighted distribution of the total 120A load current is applied, resulting in module 1 bearing approximately 73.4A and module 3 bearing only approximately 46.6A. This adjustment logic deeply integrates communication service requirements, transient voltage support, and physical thermal boundaries. Its beneficial effects are that the impedance adjustment at the front end suppresses excessive voltage drops on the bus, while the thermoelectric coupling distribution at the rear end achieves asymmetrical load balancing based on real-time physical health conditions. This avoids thermal runaway caused by prolonged heavy loads in modules with poor heat dissipation, thus improving overall reliability and service life under complex operating conditions.

[0028] In summary, the entire data processing method for power load balancing in communication base stations firstly involves constructing a feedforward model by introducing baseband-side data scheduling parameters and energy consumption benchmarks. This changes the existing feedback method where the power supply waits for the bus voltage drop before responding, enabling the quantification of transient current steps and determination of the initial jump point in advance, thus compensating for the sampling delay in the feedback loop. Furthermore, the derivative characteristics of the bus electrical state sequence are used to determine the minimum point of the bus voltage drop. This physical anchor point divides the response window into a pre-response stage supported by capacitor discharge and a post-response stage supported by inductor current takeover, avoiding the misalignment problem caused by directly scaling on the time axis. This ensures that the stage segmentation boundary aligns with the electrical evolution process. Furthermore, it enables independent evaluation of the deviation in the voltage drop slope in the first stage and the thermal resistance operating parameters in the second stage, providing data for subsequent adjustments. Furthermore, in the pre-response stage, by dynamically reducing the virtual output impedance parameter based on the step value, the rate of change of the transient inductor current is improved, suppressing the voltage drop amplitude of the bus. In the post-response stage, a current thermal tolerance margin model is established based on the real-time junction temperature, thermal resistance, and conduction loss of each module, allowing modules with good heat dissipation and high temperature margins to share more load, achieving adaptive balance in heat distribution and reducing the risk of thermal runaway caused by heat accumulation in local modules.

[0029] like Figure 2 As shown, in one specific implementation, S1 includes: S11, acquiring burst data increments as data scheduling parameters and acquiring single-bit dynamic energy consumption as an energy consumption benchmark; acquiring the current bus voltage and scheduling duration; acquiring the transient current step based on the single-bit dynamic energy consumption, burst data increments, current bus voltage, and scheduling duration. The calculation process follows the following equation:

[0030] In the above formula, It represents the transient current step, and the unit is ampere (A). This represents the energy consumption benchmark, i.e., the dynamic energy consumption per bit, measured in joules per bit (J / bit). This represents the data scheduling parameter, i.e., the burst data increment, and is measured in bits. This represents the current bus voltage, in volts (V). This represents the duration of the scheduling, in seconds (s).

[0031] S12. Take the expected moment of generating the transient current step as the starting point of the first response time domain window; obtain the current trajectory range where the actual current change occurs based on the actual output current; and establish the current trajectory range as the second response time domain window.

[0032] In this embodiment, it should be noted that in S11, transient requirements are quantized by parsing the media access control layer data. Specifically, the expression used for feedforward quantization of transient current requirements is as follows: The calculation logic of this expression is based on fundamental electromagnetism and the law of conservation of energy, aiming to directly convert the data scheduling behavior of the digital baseband domain into the physical energy demand of the analog power domain.

[0033] In the numerator of the expression, the single-bit dynamic energy consumption is multiplied by the burst data increment. The physical meaning of this multiplication is to calculate the total dynamic energy consumed by the RF unit to convert this data into electromagnetic waves for radiation within a given scheduling duration. Substituting data from a specific application scenario, when the energy consumption benchmark, i.e., the single-bit dynamic energy consumption, is... Furthermore, the data scheduling parameters in the media access control layer's data transmission queue, i.e., the burst data increment, have reached [a certain value]. When the two are multiplied, the total energy required is: .

[0034] In the denominator of the expression, the current bus voltage is multiplied by the dispatch duration, according to the power formula. It can be seen that the product of voltage and time represents the energy that the power supply can deliver within the corresponding time window for each ampere of current provided. Substituting the scenario data, the current bus voltage is... The scheduling duration is Multiplying the two together gives the value of the denominator. Its physical dimension is equivalent to joules per ampere.

[0035] Then, the total energy demand represented by the numerator is divided by the unit current energy transfer rate represented by the denominator, i.e. Divide by Calculations show that in the next... The transient current step that the internal power supply must instantaneously increase is: The fundamental reason for adopting this division operation is to break the lag limitation of passively relying on bus voltage drop sampling in the existing power supply closed-loop control. Because the existing feedback mechanism must wait for the energy of the filter capacitor to be substantially consumed, resulting in a voltage drop, before triggering the rise of the inductor current, this millisecond-level physical delay can cause low-voltage power outage risks when facing the burst traffic of 5G base stations with large bandwidth. This expression, by analyzing the incremental data packets at the communication layer, calculates the required transient step current index in advance before it is converted into the RF power at the physical layer. This allows the power supply's underlying drive controller to use this value as a basis to simultaneously lower the virtual impedance and actively increase the inductor current while transmitting data. This achieves cross-domain feedforward control of data flow predicting energy flow, effectively avoiding the phase lag problem in the feedback loop and ensuring the transient voltage support capability in the early stages of severe load changes.

[0036] In S12, a time correspondence between theoretical requirements and actual output is established to address potential time misalignments between theoretical predictions and physical execution. The expected moment of generating the transient current step is used as the starting point of the first response time-domain window; the current trajectory interval where the actual current change occurs is obtained based on the actual output current; and this current trajectory interval is established as the second response time-domain window. In the aforementioned scenario, when the calculated... After detecting the transient step current, the expected sudden change point is designated as the starting point of the first response time-domain window. Simultaneously, the actual output current of the parallel modules is continuously sampled, and the actual current trajectory range where the current surge occurs is established as the second response time-domain window. This logic ensures the alignment of the feedforward warning time with the actual hardware response time in subsequent analysis. Its beneficial effect is that it provides a reliable time-domain basis for subsequent evaluation of transient response capabilities, avoiding data misalignment issues caused by relying on a single clock source.

[0037] like Figure 3 As shown, in one specific embodiment, S2 includes: S21, obtaining the first and second derivatives of the bus electrical state sequence as a function of time; obtaining the physical moment when the first derivative is zero and the second derivative is greater than zero; and calibrating the physical moment as the minimum point of bus voltage drop indicating the end of capacitor discharge support.

[0038] S22. Define the time interval from the initial jump point to the minimum value of the bus voltage drop as the first pre-response stage and the second pre-response stage; define the time interval from the minimum value of the bus voltage drop to the moment when the current output tends to stabilize as the first post-response stage and the second post-response stage.

[0039] In this embodiment, it should be noted that in S21, the state transition anchor point in the response process is found through actual changes in electrical physical parameters, replacing the existing fixed-time-length segmentation method. The first and second derivatives of the bus electrical state sequence over time are obtained; the physical moment when the first derivative is zero and the second derivative is greater than zero is obtained; and this physical moment is marked as the minimum point of the bus voltage drop indicating the end of capacitor discharge support. Specifically, since the bus electrical state sequence is discrete sampled data, the specific process for obtaining the first and second derivatives is as follows: the bus voltage sample value of the current sampling period is subtracted from the bus voltage sample value of the previous period, and the difference is divided by the time length of the sampling period to calculate the first derivative of the current period; further, the first derivative of the current period is subtracted from the first derivative of the previous period, and the difference is again divided by the time length of the sampling period to calculate the second derivative. When the system detects that the value of the first derivative of the current cycle changes from negative to zero or positive, and the value of the second derivative is greater than zero, it locates and marks the corresponding physical moment as the minimum point of the bus voltage drop indicating the end of capacitor discharge support.

[0040] Furthermore, in Under current surge, the bus voltage from The voltage begins to drop; calculate the derivative change of this process. As the voltage drops... At the moment the voltage is about to rebound, the first derivative is zero and the second derivative is greater than zero, which is identified as the minimum point of the bus voltage drop. The logic lies in using the characteristics of calculus to capture the point where the filter capacitor stops discharging and the inductor current takes over the load. The beneficial effect of this step is that it establishes a physical dividing point that is strictly based on the laws of electrical evolution, eliminating the uncertainty brought about by artificially setting empirical time thresholds.

[0041] In S22, the entire dynamic response process is decoupled into independent stages with different physical objectives, addressing the problem of existing solutions confusing transient support with steady-state current sharing. The time interval from the initial jump point to the minimum point of the bus voltage drop is defined as the first pre-response stage and the second pre-response stage; the time interval from the minimum point of the bus voltage drop to the moment when the current output tends to stabilize is defined as the first post-response stage and the second post-response stage. In the aforementioned scenario, from the initial jump point of the command issuance to the voltage drop... The time period is divided into the pre-response phase, during which the physical behavior is capacitor discharge; while from The period from the initial rise in current until it stabilizes is designated as the post-response phase, where the physical behavior is inductive current sharing. The beneficial effect is that it reduces the continuous power supply response process to two phases with independent physical characteristics, allowing subsequent parameter evaluation and control strategies to be implemented independently for the electrical characteristics of each phase.

[0042] like Figure 4 As shown, in one specific embodiment, S3 includes: S31, by linearly fitting the voltage trajectories of the first pre-response stage (demand side) and the second pre-response stage (actual output side), extracting the reference voltage drop slope and the actual voltage drop slope respectively, calculating the absolute deviation between the two as the pre-response slope difference, and determining whether it is within the preset tolerance threshold.

[0043] S32. Real-time acquisition of the current junction temperature, ambient temperature, and instantaneous output current value of each load distribution unit (i.e., each parallel power supply submodule) is used to extract the physical characteristics required for subsequent distribution.

[0044] In this embodiment, it should be noted that in S31, within the defined pre-response stage, the deviation of the power supply's transient voltage support capability is quantitatively evaluated, addressing the issue of how to measure whether the transient response meets the standard. By fitting the voltage trajectories of the demand side and the actual output side, the reference voltage drop slope of the first pre-response stage and the actual voltage drop slope of the second pre-response stage are obtained respectively. Based on the reference voltage drop slope and the actual voltage drop slope, the pre-response slope difference is obtained. The specific process for obtaining the pre-response slope difference is as follows: the absolute value of the actual voltage drop slope of the second pre-response stage and the absolute value of the reference voltage drop slope of the first pre-response stage are extracted respectively, and the absolute value of the actual voltage drop slope is subtracted from the absolute value of the reference voltage drop slope. The resulting value is the pre-response slope difference.

[0045] The reference voltage drop slope in the first pre-response stage is derived from the minimum safe operating voltage boundary of the base station RF equipment and the nominal discharge characteristics of the bus capacitor. Specifically, the method for determining this slope is as follows: The standard discharge trajectory of the same batch of power supply equipment under rated current step load during factory testing is extracted, and combined with historical test data under a limited number of extreme burst conditions, the average slope expectation is obtained through linear fitting using the least squares method. For example, if the equipment specification requires that the 48V bus, under full-load impact, must have an allowable maximum voltage drop within 1.5V to prevent triggering the low-voltage power-off protection of the subsequent RF unit, the corresponding physical time window for capacitor safe discharge support is typically 375μs. Dividing this set of historically measured extreme boundary values ​​by (1.5V / 0.375ms) yields a reference voltage drop slope of -4V / ms, which serves as a benchmark for assessing whether transient support capability meets the standard.

[0046] Depending on the scenario, if the baseline slope is The actual observed slope is The difference between the two and the absolute value is the difference in the slope of the previous response. Then determine whether it exceeds the preset limit. Tolerance threshold. The logic behind calculating the difference is to intuitively reflect the degree of deviation between the actual hardware response speed and baseband requirements. The beneficial effect of this step is to provide a quantitative transient assessment index, which can promptly detect low-voltage risks when dealing with sudden loads, and provide a basis for determining the conditions for triggering subsequent impedance adjustment.

[0047] In S32, after establishing the post-response phase, the physical parameters required for subsequent steady-state heat distribution are acquired in real time, addressing the issue that existing current sharing methods do not consider actual operating environment conditions. In the second post-response phase, the current junction temperature, ambient temperature, and instantaneous output current value of each load distribution unit are acquired in real time to extract the physical characteristics required for subsequent distribution. Specifically, the actual operating status of each parallel module is continuously monitored; for example, the current junction temperature of the first module is read as... And the current junction temperature of module three is The logic behind this step is that the heat dissipation capacity of a power module is not statically constant, but is affected by dynamic factors such as actual heat accumulation and ventilation. Only by obtaining real-time physical parameters can its true operating state be reflected. Its beneficial effect is that it provides multi-dimensional source data support for establishing a dynamic thermal-electric coupling model, overcoming the limitations of existing methods that rely solely on fixed factory parameters.

[0048] like Figure 5 As shown, in one specific embodiment, S4 includes: S41, calculating the impedance compensation coefficient based on the difference between the transient current step and the previous response slope; reducing the virtual output impedance parameter in the control loop of each load distribution unit in the power supply state based on the impedance compensation coefficient, so as to globally improve the rate of change of transient inductor current over time. By systematically reducing the virtual output impedance parameter, the rise rate of inductor current can be improved overall, making up for the energy gap during the discharge of the bus capacitor.

[0049] S42. Obtain the current junction temperature and the safe junction temperature threshold; obtain the temperature margin based on the safe junction temperature threshold and the current junction temperature; obtain the current thermal capacity margin for each load distribution unit based on the temperature margin, thermal resistance characteristic parameters, and conduction loss parameters. The calculation process follows the following equation:

[0050] In the above formula, Represents the load distribution unit The current thermal capacity margin, expressed in amperes squared ( ); This represents the safe junction temperature threshold set by the load distribution unit, in degrees Celsius (°C). Represents the load distribution unit The current junction temperature, in degrees Celsius (°C); Represents the load distribution unit Thermal resistance characteristics, in degrees Celsius per watt (°C / W); Represents the load distribution unit The equivalent on-resistance of an internal switching device, i.e., the conduction loss parameter, is expressed in ohms (Ω). ).

[0051] Among them, the safe junction temperature threshold It is determined based on the datasheet specifications (SOA curves) of the power semiconductor devices within the load sharing unit and after applying an industrial-grade reliability derating factor. The method for determining the value involves extracting the maximum rated junction temperature of the device under extreme operating conditions and applying it according to the complexity of the outdoor operating environment of the base station. to The derating factor. For example, the rated limiting junction temperature of the selected power MOSFET is... Considering the extreme conditions of high-temperature seasons and filter aging, the application... A derating factor of multiples is used for safety limits, thereby calibrating the safe junction temperature threshold to... This serves as the physical boundary for calculating the thermal capacity margin, effectively extending the service life of power devices.

[0052] The thermal resistance and conduction loss parameters are not static constants, but are obtained by establishing a physical characteristic map of the equipment and combining it with periodic state estimation and updates. Specifically, the system writes the calibration baseline values ​​under factory testing conditions during the initialization phase, and periodically executes self-test routines during periods of low business activity. By injecting a constant small test current and continuously collecting historical steady-state temperature rise data, the system recalibrates the current true thermal resistance and conduction resistance using a linear regression model of temperature rise and corresponding power loss. For example, when the system performs a constant current self-test on module 3 at night under low load, it finds that under the same output power consumption, this module's steady-state temperature rise is 50% higher than the factory historical record due to severe dust accumulation on the filter caused by long-term operation. The system then automatically corrects the model parameters using a linear regression algorithm, automatically updating the module's thermal resistance parameter from the factory preset 0.6℃ / W to the current 0.9℃ / W, ensuring that the underlying thermoelectric coupling parameters accurately reflect the aging and heat dissipation status of the hardware.

[0053] S43. Calculate the linear current margin by taking the square root of the current thermal tolerance margin for each load distribution unit, and then determine the distribution weight based on this linear current margin. Finally, weight the total load current according to the distribution weight to obtain the target output current for the corresponding load distribution unit. The calculation process follows the equation below:

[0054] In the above formula, Representatives are assigned to the load distribution unit. The target output current, in amperes (A); This represents the total load current required by the communication base station, measured in amperes (A). Represents the load distribution unit The current thermal capacity margin, expressed in amperes squared ( ); Representing the The current thermal capacity margin of each unit, expressed in amperes squared ( ); This represents the total number of load distribution units connected in parallel.

[0055] Using the above method, when a sudden service arrives, the virtual impedance of the front-end response can be adjusted in advance using baseband scheduling parameters to protect the bus voltage. In the subsequent back-end process, the current can be rebalanced based on the real-time heat dissipation capacity and temperature margin of each unit to avoid operational risks caused by overheating of individual units.

[0056] In this embodiment, it should be noted that in S41, for the transient deviation discovered in S31, the physical response trajectory of the preceding stage is reshaped by adjusting the control parameters to solve the problem of excessive bus voltage drop. If the difference in the preceding response slope exceeds the tolerance threshold, the virtual output impedance parameter of the second preceding response stage is adjusted according to the difference in the preceding response slope and the transient current step. In specific operation, when it is determined that... The slope difference exceeds At the threshold, the impedance compensation coefficient is calculated based on the transient demand of 25A, and the virtual output impedance parameter in the control loop of all parallel load distribution units participating in power supply is synchronously reduced from the initial 10mΩ to 6mΩ.

[0057] Specifically, the calculation process follows the following equation: ; ;in, Represents the impedance compensation factor, in units of ; This represents the preset system impedance adjustment gain constant, in units of... ; This represents the difference in slope between the previous and current responses; It represents the transient current step, and its unit is A; and These represent the virtual output impedance parameters before and after adjustment, respectively. Among them, the system impedance adjustment gain constant... This is achieved by establishing a closed-loop small-signal mathematical model of the power module and combining empirical balance parameters determined through multiple iterative simulation experiments. The core objective is to ensure that the impedance compensation is sufficient to offset voltage drops while preventing oscillations and instability in the control loop due to excessive gain. The value selection process is as follows: Simulate a 25A current step in the simulation environment, and adjust the gain value from... Scanning was performed by gradually increasing the step size; comparison revealed that when the value was [value missing]... At that time, the system can generate approximately The precise compensation amount raised the minimum voltage drop point of the bus from 47.1V to above 47.5V, and there was no obvious overshoot during the adjustment process, achieving the optimal solution for response speed and stability.

[0058] The logic is that reducing the virtual output impedance parameter can increase the rate of change of the internal inductor current over time, thereby increasing the energy replenishment during the discharge of the bus capacitor. The beneficial effect of this step is that it interferes with the underlying hardware response characteristics, raising the minimum voltage drop point under similar sudden load conditions to a higher level. The above measures ensure the operating voltage status of the internal components of the base station.

[0059] In S42, after entering steady state, the thermal tolerance of each module is evaluated through physical parameter modeling, addressing the issue of existing current sharing methods that do not assess the risk of local overheating. The current junction temperature and a safe junction temperature threshold are obtained; a temperature margin is derived based on the safe junction temperature threshold and the current junction temperature; and the current thermal capacity margin for each load distribution unit is obtained based on the temperature margin, thermal resistance parameters, and conduction loss parameters. Specifically, the expression used to quantify the steady-state thermal load state of each power supply submodule is as follows: The calculation logic of this expression is to strictly convert the thermodynamic boundary constraints of semiconductor devices into available tolerance parameters in the electrical control loop, thereby establishing allocation constraints based on physical health in the parallel connection of multiple modules.

[0060] The numerator of the expression calculates the absolute temperature margin remaining before the module reaches the physical damage boundary by subtracting the current junction temperature collected in real time by the temperature sensor from the set safe junction temperature threshold. Taking the application scenario as an example, the set safe junction temperature threshold for the first power supply submodule is... The current junction temperature is The calculated temperature margin is as follows: The third power supply submodule, whose heat dissipation has deteriorated due to dust accumulation on the filter, currently has a junction temperature as high as [missing information]. The calculated temperature margin is only .

[0061] The denominator of the expression is the product of the thermal resistance parameter and the conduction loss parameter of the switching device, based on the fundamental heat conduction formula. and the formula for conduction loss of electrical power The mapping relationship between temperature rise and current can be derived as follows: Therefore, by dividing the temperature margin of the numerator by the product of the thermal resistance and the electrical resistance of the denominator, we can inversely solve for the square of the steady-state current increase allowed by the module while maintaining the current heat dissipation level.

[0062] Substituting the scenario data, the thermal resistance characteristic parameters of the first module are: The conduction loss parameter is The product of the denominators is , using molecules Dividing by this value, the calculated current thermal capacity margin of the first module is approximately Similarly, the thermal resistance parameter of module three increased due to impeded heat dissipation. The conduction loss parameters are the same. The product of the denominators is , using molecules Dividing by this value, the calculated current thermal capacity margin for the third module is approximately... This calculation process, which uses the square term of the current for margin assessment, truly reflects the objective physical law that the conduction loss inside power electronic devices increases non-linearly with the output current in a quadratic manner. It solves the problem of forced current sharing caused by assuming that the physical states of all parallel modules are completely identical in existing current sharing strategies. By transforming the complex real-time thermodynamic environment into controllable electrical parameters, it provides objective data support for subsequent asymmetric load balancing and avoids issuing excessive current commands to modules with deteriorating heat dissipation conditions.

[0063] In S43, the total current is redistributed based on the physical margin calculated in S42 to address the uneven heat distribution problem during multi-module parallel operation. The square root of the current thermal tolerance margin of each load allocation unit is taken to obtain the corresponding linear current margin. The allocation weight is obtained based on the ratio of the linear current margin of each load allocation unit to the sum of the linear current margins of all load allocation units. The total load current is then weighted and allocated according to the allocation weight to obtain the target output current for the corresponding load allocation unit. Specifically, the expression used to generate the final target allocation instruction for each power supply submodule is as follows: This expression completes the closed-loop transformation from the physical tolerance assessment in the preceding steps to the final steady-state equilibrium control command.

[0064] Furthermore, since the current thermal capacity limit calculated in S42 has the dimension of ampere squares, the expression first performs a square root operation on the current thermal capacity limit of each module to restore its dimension to the linear current dimension of the first power, ensuring that it matches the actual physical current in subsequent proportional calculations. Combining the aforementioned scenario data, the current thermal capacity limit of the first module is... After taking the square root, the linear current margin is approximately Current thermal capacity limit of module 3 After taking the square root, the linear current margin is approximately .

[0065] Furthermore, the denominator of the expression sums the linear current margins of all parallel load distribution units to calculate the total physical support margin that the entire power supply can currently provide within the safe temperature boundary, i.e. The fractional approach, by dividing the linear margin of a single module by the total margin, quantifies the relative weight that the module should bear based on the current hardware health state. The calculated weight of the first module is approximately... The weight allocated to module number three is approximately .

[0066] Finally, multiply the required total load current by the weighted distribution of each module to obtain the specific current output required by the corresponding module in the steady-state phase. When the total load current required by the communication base station is... At that time, the target output current of the first module is The target output current of the third module is By adopting this weighted proportional calculation process with a normalized denominator, on the one hand, the sum of the target output current allocated to all modules is dynamically matched with the total load current required for base station service operation, maintaining the power supply and demand balance. On the other hand, it substantially solves the problem of local heat accumulation in the long-term operation of multiple modules. By guiding more current load to healthy modules with high heat dissipation efficiency and large temperature margin, it reduces the conduction loss and thermal stress of deteriorated modules, and realizes a level-adaptive thermal distribution balance based on underlying physical parameters.

[0067] This invention also provides a communication base station power load balancing data processing system, which is used to implement a communication base station power load balancing data processing method. The system includes: The parameter acquisition and window establishment module is used to acquire data scheduling parameters and energy consumption benchmarks, and acquire transient current step based on the data scheduling parameters and energy consumption benchmarks; determine the starting jump point of the first response time domain window based on the transient current step, acquire the actual output current, and determine the second response time domain window based on the actual output current; The physical segmentation module is used to obtain the bus electrical state sequence and obtain the minimum value point of bus voltage drop based on the bus electrical state sequence; based on the minimum value point of bus voltage drop, the first response time domain window is divided into a first pre-response stage and a first post-response stage, and the second response time domain window is divided into a second pre-response stage and a second post-response stage. The deviation assessment module is used to obtain the reference voltage drop slope in the first pre-response stage and the actual voltage drop slope in the second pre-response stage, and to obtain the difference in pre-response slope based on the reference voltage drop slope and the actual voltage drop slope. The step reduction adjustment module is used to adjust the virtual output impedance parameters of the second front response stage according to the front response slope difference and transient current step if the front response slope difference exceeds the tolerance threshold. It obtains the thermal resistance characteristic parameters and conduction loss parameters of the load distribution unit, obtains the current thermal tolerance margin according to the thermal resistance characteristic parameters and conduction loss parameters, and distributes the total load current according to the current thermal tolerance margin to generate the target output current.

[0068] In one specific implementation, the parameter acquisition and window establishment module is also used to: acquire burst data increments as data scheduling parameters and acquire single-bit dynamic energy consumption as energy consumption benchmarks; acquire the current bus voltage and scheduling duration; and acquire transient current step based on single-bit dynamic energy consumption, burst data increments, current bus voltage, and scheduling duration.

[0069] In one specific implementation, the order reduction adjustment module is further configured to: obtain the current junction temperature and the safe junction temperature threshold; obtain the temperature margin based on the safe junction temperature threshold and the current junction temperature; obtain the current thermal tolerance margin of each load distribution unit based on the temperature margin, thermal resistance characteristic parameters, and conduction loss parameters; perform a square root operation on the current thermal tolerance margin of each load distribution unit to obtain the corresponding linear current margin; obtain the allocation weight based on the ratio of the linear current margin of each load distribution unit to the sum of the linear current margins of all load distribution units; and perform weighted allocation of the total load current based on the allocation weight to obtain the target output current of the corresponding load distribution unit.

[0070] To further clarify the operating mechanism and physical quantification process of the technical solution of the present invention in actual industrial production lines, the following analysis will be conducted in detail on the underlying derivation logic of the power load balancing data processing method and system of communication base stations, taking into account a scenario containing specific parameters and data.

[0071] In practical application scenarios, taking a high-load 5G macro base station in the core area of ​​a city as an example, the base station adopts a power supply system composed of four power modules connected in parallel, with a rated DC output voltage of... for During the morning rush hour at 8:00 AM, the baseband processing unit (BBU) predicts the next radio scheduling frame through the media access control layer. for (Right now Furthermore, due to a large number of users connecting, there is a sudden increase in data in the data sending queue. Reached At this point, the system enters S11, based on the current dynamic power consumption per bit of the RF unit. for (Right now Perform feedforward calculations based on the transient current demand feedforward model formula. Substituting the data, we can calculate... This means that in the next... Internally, the power system needs to be increased instantaneously. With the current support, the calculation result directly establishes the starting jump point of the first response time domain window.

[0072] After entering S21, the system monitors the bus voltage in real time. Fluctuations. Due to a sudden increase in load, the bus voltage... The voltage begins to drop rapidly, and the system calculates the first derivative of the voltage. from It becomes a negative value. The drop lasts approximately... Then, the voltage dropped to And it begins to recover due to feedback from the power module, at which point the system captures the first derivative as And the second derivative is greater than The physical moment is calibrated as the minimum point of the bus voltage drop. This physical anchor point divides the window into a pre-response stage ("capacitor support segment") and a post-response stage ("inductor current ramp-up segment"). At this point, executing S31, the system compares and finds that the preset reference voltage drop slope should be maintained at... The slope of the second pre-response phase actually observed is The absolute deviation is It exceeded the preset limit. The tolerance threshold is used to determine whether the system has an excessively large risk of transient voltage drops.

[0073] In response to the aforementioned deviation, the system immediately executes S41. The controller, based on... The step demand dynamically reduces the first response lag. No. and The virtual output impedance parameters in the voltage loop of the power module. Simultaneously, the system's preset impedance adjustment gain constant. for Substitute It is 0.5V / ms (equivalent absolute value). It is 25A, calculated as follows Therefore, the virtual output impedance is reduced from 10mΩ to 6mΩ; by adjusting the virtual impedance from the initial... Reduce to This causes the rate of change of current in the power inductor to... This significantly improves the energy supply to the bus capacitor in a very short time. This adjustment effectively alters the voltage drop trajectory during the initial response phase, raising the minimum voltage drop point under subsequent similar fluctuations to a higher level. The above ensures that sensitive radio frequency components inside the base station will not generate bit errors or restart due to instantaneous low voltage, demonstrating the role of baseband data feedforward and physical boundary segmentation in ensuring transient stability.

[0074] The system then enters S42 to handle the steady-state current distribution problem in the post-response phase, at which point the thermal load of each module needs to be considered. (The sentence is incomplete and ends abruptly.) Taking module number 1 as an example, its power devices have a maximum safe junction temperature. for The temperature sensor measures the current junction temperature. for The overall thermal resistance of the heat dissipation structure for On-loss parameters of internal switching transistors for (Right now According to the formula for the thermal capacity limit of the current square. Calculated This value quantifies the physical margin of the steady-state current square value that the module can withstand under current heat dissipation conditions without the risk of overheating.

[0075] At the same time, the The heat dissipation of module 1 has deteriorated due to dust accumulation on the fan filter, and its thermal resistance has decreased. Rise to And the current junction temperature Relatively high, reaching Substitute into the same formula to calculate the first... The margin of module number: As can be seen from the comparison, although the rated power of each module is the same, the... The current squared thermal capacity margin of module number 1 is much larger than that of module number 2. Module No. 1. This step truly reflects the energy mapping relationship between junction temperature, thermal resistance, and conduction losses in a power electronic system, providing an objective physical basis for subsequent asymmetric current distribution.

[0076] Finally, execute S43, assuming the base station's total load current at this time... Stable at To simplify the explanation, if we only consider the first... No. and Module No. 1 has current squared thermal capacity margins of 1 / 2 and 1 / 2 respectively. and According to the target current allocation instruction formula First, calculate the square root of each remainder: , Then the first The target output current allocated to module number And the first The current allocated to module number .

[0077] As can be seen from the complete calculation logic above, the system does not adopt an average current distribution approach, but rather uses thermal-electric coupling reconfiguration to allow modules with better heat dissipation and lower temperatures to bear approximately [amount missing]. The heavy load, while allowing modules with a higher risk of overheating to bear only about [amount missing]. This approach avoids the load of existing flow sharing strategies. The module may experience continuous overheating leading to protection tripping or insulation aging risks. By combining front-end virtual impedance adjustment with back-end current distribution of thermal margin, this technical solution effectively balances the physical health of the parallel power supply module group while ensuring base station communication quality, achieving reliable operation of the power electronic system under variable load conditions.

[0078] The preferred embodiments of this disclosure have been described in detail above with reference to the accompanying drawings. However, this disclosure is not limited to the specific details of the above embodiments. Within the scope of the technical concept of this disclosure, various simple modifications can be made to the technical solutions of this disclosure, and these simple modifications all fall within the protection scope of this disclosure.

[0079] It should also be noted that the various specific technical features described in the above embodiments can be combined in any suitable manner without contradiction. To avoid unnecessary repetition, this disclosure will not describe the various possible combinations separately.

[0080] Furthermore, various different embodiments of this disclosure can be combined in any way, as long as they do not violate the spirit of this disclosure, they should also be regarded as the content disclosed in this disclosure.

[0081] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be covered within the scope of the claims and specification of the present invention.

Claims

1. A data processing method for power load balancing in a communication base station, characterized in that, The methods include: Obtain data scheduling parameters and energy consumption benchmarks, and obtain transient current step values ​​based on data scheduling parameters and energy consumption benchmarks; The starting point of the first response time-domain window is determined based on the transient current step, the actual output current is obtained, and the second response time-domain window is determined based on the actual output current. Obtain the bus electrical state sequence, and obtain the minimum point of bus voltage drop based on the bus electrical state sequence; Based on the minimum point of bus voltage drop, the first response time domain window is divided into a first pre-response stage and a first post-response stage, and the second response time domain window is divided into a second pre-response stage and a second post-response stage. The reference voltage drop slope in the first pre-response stage and the actual voltage drop slope in the second pre-response stage are obtained respectively, and the difference in pre-response slope is obtained based on the reference voltage drop slope and the actual voltage drop slope. If the difference in the slope of the preceding response exceeds the tolerance threshold, the virtual output impedance parameter of the second preceding response stage is adjusted according to the difference in the slope of the preceding response and the transient current step, the thermal resistance characteristic parameter and conduction loss parameter of the load distribution unit are obtained, and the current thermal tolerance margin is obtained according to the thermal resistance characteristic parameter and conduction loss parameter. The total load current is distributed according to the current thermal tolerance margin to generate the target output current.

2. The communication base station power load balancing data processing method according to claim 1, characterized in that, The process of acquiring data scheduling parameters and energy consumption benchmarks, and acquiring transient current step values ​​based on these parameters and benchmarks, includes: Acquire burst data increments as data scheduling parameters, and obtain single-bit dynamic energy consumption as energy consumption benchmark; Get the current bus voltage and scheduling duration; The transient current step is obtained based on the single-bit dynamic energy consumption, burst data increment, current bus voltage, and scheduling duration.

3. The communication base station power load balancing data processing method according to claim 2, characterized in that, The step of determining the starting transition point of the first response time-domain window based on the transient current step, obtaining the actual output current, and determining the second response time-domain window based on the actual output current includes: The expected moment when the transient current step is generated is taken as the starting point of the first response time domain window; The current trajectory range in which the actual current change occurs is obtained based on the actual output current; The current trajectory range is defined as the second response time domain window.

4. The communication base station power load balancing data processing method according to claim 1, characterized in that, The step of obtaining the bus electrical state sequence and obtaining the minimum point of bus voltage drop based on the bus electrical state sequence includes: Obtain the first and second derivatives of the bus electrical state sequence as a function of time; Obtain the physical moment when the first derivative is zero and the second derivative is greater than zero; The physical moment is calibrated as the minimum point of bus voltage drop at which the capacitor discharge support ends.

5. The communication base station power load balancing data processing method according to claim 4, characterized in that, The first response time-domain window is divided into a first pre-response stage and a first post-response stage based on the minimum point of the bus voltage drop, and the second response time-domain window is divided into a second pre-response stage and a second post-response stage, including: The time interval from the initial jump point to the minimum value of the bus voltage drop is defined as the first pre-response stage and the second pre-response stage. The time interval from when the bus voltage drops to its minimum point to when the current output tends to stabilize is defined as the first post-response stage and the second post-response stage.

6. The communication base station power load balancing data processing method according to claim 1, characterized in that, The adjustment of the virtual output impedance parameters in the second pre-response stage based on the difference in pre-response slope and the transient current step includes: The impedance compensation coefficient is calculated based on the difference between the transient current step and the slope of the previous response. The virtual output impedance parameter in the control loop of each load distribution unit under power supply state is reduced by reducing the impedance compensation coefficient, so as to globally improve the rate of change of transient inductor current over time.

7. The communication base station power load balancing data processing method according to claim 1, characterized in that, The process of obtaining the thermal resistance characteristic parameters and conduction loss parameters of the load distribution unit, obtaining the current thermal capacity margin based on the thermal resistance characteristic parameters and conduction loss parameters, and distributing the total load current according to the current thermal capacity margin to generate the target output current includes: Obtain the current junction temperature and the safe junction temperature threshold; The temperature margin is obtained based on the safe junction temperature threshold and the current junction temperature; Based on the temperature margin, thermal resistance characteristic parameters, and conduction loss parameters, obtain the current thermal capacity margin for each load distribution unit. The allocation weight is obtained based on the current thermal capacity margin of each load allocation unit; The total load current is weighted and allocated according to the allocation weights to obtain the target output current of the corresponding load allocation unit.

8. A power load balancing data processing system for a communication base station, characterized in that, include: The parameter acquisition and window establishment module is used to acquire data scheduling parameters and energy consumption benchmarks, and to acquire transient current step values ​​based on the data scheduling parameters and energy consumption benchmarks. The starting point of the first response time-domain window is determined based on the transient current step, the actual output current is obtained, and the second response time-domain window is determined based on the actual output current. The physical segmentation module is used to obtain the bus electrical state sequence and obtain the minimum value point of bus voltage drop based on the bus electrical state sequence; based on the minimum value point of bus voltage drop, the first response time domain window is divided into a first pre-response stage and a first post-response stage, and the second response time domain window is divided into a second pre-response stage and a second post-response stage. The deviation assessment module is used to obtain the reference voltage drop slope in the first pre-response stage and the actual voltage drop slope in the second pre-response stage, and to obtain the difference in pre-response slope based on the reference voltage drop slope and the actual voltage drop slope. The step reduction adjustment module is used to adjust the virtual output impedance parameters of the second front response stage according to the front response slope difference and transient current step if the front response slope difference exceeds the tolerance threshold. It obtains the thermal resistance characteristic parameters and conduction loss parameters of the load distribution unit, obtains the current thermal tolerance margin according to the thermal resistance characteristic parameters and conduction loss parameters, and distributes the total load current according to the current thermal tolerance margin to generate the target output current.

9. The communication base station power load balancing data processing system according to claim 8, characterized in that, The parameter acquisition and window establishment module is also used for: Acquire burst data increments as data scheduling parameters, and obtain single-bit dynamic energy consumption as energy consumption benchmark; Get the current bus voltage and scheduling duration; The transient current step is obtained based on the single-bit dynamic energy consumption, burst data increment, current bus voltage, and scheduling duration.

10. The communication base station power load balancing data processing system according to claim 8, characterized in that, The order reduction adjustment module is also used for: Obtain the current junction temperature and the safe junction temperature threshold; The temperature margin is obtained based on the safe junction temperature threshold and the current junction temperature; Based on the temperature margin, thermal resistance characteristic parameters, and conduction loss parameters, obtain the current thermal capacity margin for each load distribution unit. The allocation weight is obtained based on the current thermal capacity margin of each load allocation unit; The total load current is weighted and allocated according to the allocation weights to obtain the target output current of the corresponding load allocation unit.

Images