Regulation and control method, device and equipment of load power supply loop, medium and program product

Abstract

The invention discloses a regulation and control method and device for a load power supply loop, equipment, a medium and a program product, and relates to the technical field of servers, and the method comprises the steps: obtaining the current change rate, the contact temperature rise rate and the voltage change rate of a load interface when a hot plug event occurs, so as to comprehensively reflect the rapid change condition of a circuit state at the hot plug moment; then, inputting the data into a surge probability prediction model to obtain a surge occurrence probability so as to pre-judge a surge risk in advance; and when the surge occurrence probability is greater than or equal to a first preset probability, the resistance-capacitance parameters of the load power supply loop are calculated according to the monitored load interface current, load interface voltage and load interface temperature, so that when a hot plug event occurs, the circuit can be rapidly adjusted according to the resistance-capacitance parameters, abnormal changes of the current and the voltage are inhibited in time, and the reliability of the hot plug event is improved. The protection response speed is improved, equipment damage caused by untimely protection is effectively avoided, and the server is ensured to stably operate under complex conditions such as load hot plug.

Description

Technical Field

[0001] The present application relates to the field of server technology, and in particular to a method, device, equipment, medium and program product for controlling a load power supply circuit. Background Art

[0002] The stable operation of servers is highly dependent on continuous and effective heat dissipation support. As the core heat dissipation component of most current server systems, fans are crucial for reliable operation. To improve the convenience of server maintenance and the continuous operation of the system, the cooling fans in mainstream servers are currently generally equipped with hot-swappable functions, allowing the fans to be directly plugged in and out while the system is running. However, the relevant technology's protection measures for the fan hot-swap process are still incomplete. Typically, fans are only protected by electronic fuses (eFuses) to deal with abnormal conditions such as overcurrent, overvoltage, or short circuits, preventing damage to the fan and related circuits due to these faults.

[0003] While eFuses can provide a certain degree of short-circuit and overcurrent protection, in the specific scenario of fan hot-swapping, the rapid changes in system status can easily generate instantaneous inrush currents, which can cause the eFuse to malfunction and even damage related circuits. Furthermore, the eFuse's thermal shutdown function relies on a rise in the temperature of its internal integrated circuits to trigger, which often causes the protection response to lag behind actual needs. This is especially true during rapid hot-swapping events, and this lag can lead to untimely protection, potentially causing equipment damage and other problems. Summary of the Invention

[0004] The present application provides a method, device, equipment, medium and program product for regulating a load power supply circuit, so as to at least solve the problem in the related art that surge current generated by the load during hot plugging causes system failure.

[0005] The present application provides a method for regulating a load power supply circuit, comprising: when a baseboard management controller monitors a hot plug event of a load, obtaining the current change rate, contact heating rate and voltage change rate of the load interface; wherein the hot plug event is identified by a level signal of an in-place detection pin of the load power supply circuit; inputting the current change rate, contact heating rate and voltage change rate into a surge probability prediction model to obtain a surge occurrence probability output by the surge probability prediction model; wherein the surge probability prediction model is constructed based on a long short-term memory network and is trained based on a historical data set; when the surge occurrence probability is greater than or equal to a first preset probability, calculating the resistance and capacitance parameters of the load power supply circuit based on the monitored load interface current, load interface voltage and load interface temperature; and regulating the load power supply circuit based on the resistance and capacitance parameters.

[0006] The present application also provides a control device for a load power supply circuit, comprising: An acquisition module is used to acquire the current change rate, contact temperature rise rate, and voltage change rate of the load interface when the baseboard management controller monitors the hot plug event of the load; A surge prediction module is used to input the current change rate, contact temperature rise rate, and voltage change rate into a surge probability prediction model to obtain the surge occurrence probability output by the surge probability prediction model; the surge probability prediction model is constructed based on a long short-term memory network and trained using a historical data set; a calculation module, configured to calculate resistance and capacitance parameters of a load power supply circuit based on the monitored load interface current, load interface voltage, and load interface temperature when the probability of a surge occurring is greater than or equal to a first preset probability; The control module is used to control the load power supply circuit according to the resistance and capacitance parameters.

[0007] The present application also provides an electronic device, comprising: a memory for storing a computer program; and a processor for implementing the steps of the above-mentioned method for controlling a load power supply circuit when executing the computer program.

[0008] The present application also provides a computer-readable storage medium, in which a computer program is stored. When the computer program is executed by a processor, the steps of the above-mentioned method for controlling the load power supply circuit are implemented.

[0009] The present application also provides a computer program product, including a computer program, which implements the steps of the above-mentioned load power supply circuit control method when the computer program is executed by a processor.

[0010] When a hot plug event occurs, the present application obtains the current change rate, contact heating rate and voltage change rate of the load interface. These key indicators comprehensively reflect the rapid change of the circuit state at the moment of hot plugging; then, these key indicators are input into the surge probability prediction model to obtain the surge occurrence probability, so as to predict the surge risk in advance; when the surge occurrence probability is greater than or equal to the first preset probability, it means that there may be a large surge risk. At this time, the resistance and capacitance parameters of the load power supply circuit are calculated based on the monitored load interface current, load interface voltage and load interface temperature, so that when a hot plug event occurs, the circuit can be quickly adjusted according to the resistance and capacitance parameters, and abnormal changes in current and voltage can be suppressed in time, thereby improving the protection response speed, effectively avoiding equipment damage caused by untimely protection, and ensuring stable operation of the server under complex situations such as load hot plugging. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] In order to more clearly illustrate the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0012] Figure 1 A schematic diagram of a specific hardware architecture on which execution of a method for controlling a load power supply circuit provided in an embodiment of the present application depends; Figure 2 A schematic flow chart of a method for controlling a load power supply circuit provided in an embodiment of the present application; Figure 3 A schematic diagram of the structure of an energy feedback circuit provided in an embodiment of the present application; Figure 4 A flow chart of dynamically switching the resistance and capacitance parameters of a load power supply circuit provided in an embodiment of the present application; Figure 5 Schematic diagram of the process of controlling the load power supply circuit provided in the embodiment of the present application Figure 2 ; Figure 6 A schematic structural diagram of a control device for a load power supply circuit provided in an embodiment of the present application; Figure 7 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0013] The following will be combined with the accompanying drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0014] It should be noted that, in the description of this application, the terms "comprises," "includes," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or device. The terms "first," "second," etc., in this application are used to distinguish similar objects, and are not used to describe a particular order or sequence.

[0015] In order to more clearly illustrate the embodiments of the present application, the following briefly introduces the technical terms used in the embodiments: An electronic fuse (eFuse) is a programmable protection device based on semiconductor technology that replaces traditional mechanical fuses. By monitoring circuit parameters such as current, voltage, and temperature, it can quickly disconnect the circuit (typically with a response time of microseconds) when an anomaly such as overcurrent or overvoltage occurs. Some models also support a reset function and are reusable. They are widely used in power management, consumer electronics, and other fields.

[0016] Inrush current refers to the short-term peak current generated by the circuit at the moment of connection (or when the equipment is started), which far exceeds the steady-state operating current.

[0017] A negative temperature coefficient (NTC) refers to the physical property of a device where its resistance decreases as the temperature rises, and vice versa. Components with this property, such as NTC thermistors, are commonly used in temperature measurement, circuit current limiting (for power-on surge protection), and temperature compensation.

[0018] An analog-to-digital converter (ADC) is an electronic device that converts continuously varying analog signals (such as voltage and current) into discrete digital signals (binary code). The conversion process typically involves three core steps: sampling, quantization, and encoding. Its performance is determined by its resolution (number of bits, such as 8 or 12 bits), conversion rate, and accuracy. ADCs are the key interface between analog circuits and digital systems (such as microcontrollers and processors), and are widely used in fields such as sensor signal acquisition and audio processing.

[0019] The Long Short-Term Memory (LSTM) network is a special variant of the Recurrent Neural Network (RNN) designed to address the long-sequence dependency problem of traditional RNNs, namely their inability to effectively remember earlier information in long sequences. By designing a forget gate, input gate, and output gate structure, it can selectively retain or discard historical information. It can process time-series data such as text, speech, and time series signals, and is commonly used in tasks such as natural language processing, speech recognition, and time series prediction.

[0020] Overcurrent protection (OCP) is a circuit safety protection mechanism. Its core function is to trigger protective actions such as circuit disconnection and current limiting when the current in the circuit exceeds a preset safety threshold to prevent excessive current from damaging loads, wires, or power devices.

[0021] A MOSFET array is a device or circuit unit consisting of multiple metal-oxide-semiconductor field-effect transistors (MOSFETs) connected in a specific topology (such as parallel, series, or mixed connections). Its purpose is to improve overall current carrying capacity, increase withstand voltage, or implement complex switching logic. It is widely used in high-power scenarios such as high-power power conversion, motor control, and load switching, and can reduce power loss and heat generation in a single MOSFET.

[0022] Back EMF refers to the electromotive force (EMF) generated by a conductor or coil in electromagnetic induction, due to changes in its own current or movement in a magnetic field. This force, which is opposite to the external driving voltage, is commonly found in inductive components such as inductors, motors, and transformers. For example, when a motor is running, the back EMF generated by the coil cutting through the magnetic field suppresses the input current. If the power is suddenly lost, the back EMF can cause high voltage, requiring circuit protection devices such as freewheeling diodes.

[0023] In order to enable those skilled in the art to better understand the present application, the present application is further described in detail below with reference to the accompanying drawings and specific implementation methods.

[0024] In conjunction with the specific application environment architecture or specific hardware architecture on which the execution of the load power supply circuit control method depends, the specific application environment architecture or specific hardware architecture is described herein.

[0025] like Figure 1 As shown, Figure 1 A schematic diagram of a specific hardware architecture on which the method for controlling a load power supply circuit is implemented.

[0026] The intelligent prediction and pre-adjustment system collects current, temperature, and voltage signals from the load power supply circuit through high-frequency current sensors, temperature detection modules, and voltage detection modules. A field-programmable gate array (FPGA) predictive controller monitors the collected multi-dimensional data, analyzing whether the circuit may face risks such as surges, and generates impedance adjustment instructions to provide a basis for subsequent dynamic adjustment of circuit parameters and implement staged pre-charging.

[0027] In the dynamic RC control system, the proportional-integral-derivative (PID) control chip receives "impedance adjustment instructions" from the "intelligent prediction and pre-regulation system" and generates "RC adjustment instructions" based on its own calculations. The digitally controlled resistor array and RC network dynamically adjust the resistance and capacitance parameters of the load power supply circuit according to the RC adjustment instructions, optimizing the impedance characteristics to suppress surges and ensure stable power supply.

[0028] The energy buffering and recycling module, when there is residual surge energy in the circuit, directs the remaining surge energy into the energy storage capacitor bank through the energy feedback circuit for storage. If the system bus voltage is insufficient or the load operating parameters do not meet the standards, the energy storage capacitor bank can release energy through the energy feedback circuit to reversely supplement the load power supply circuit, realizing energy recycling and improving energy efficiency.

[0029] The specific hardware architecture on which the control method of the load power supply circuit is executed implements full-link protection of prediction, suppression and recovery, which not only enhances the stability of the load power supply, but also improves energy utilization efficiency while taking into account both safety and energy efficiency.

[0030] An embodiment of the present application provides a method for regulating a load power supply circuit. The method is described in detail in conjunction with the execution flow of the method for regulating a load power supply circuit.

[0031] like Figure 2 As shown, the control method of the load power supply circuit includes the following steps S201 to S204: S201 : When a baseboard management controller monitors a hot plug event of a load, the baseboard management controller obtains a current change rate, a contact temperature rise rate, and a voltage change rate of a load interface.

[0032] The load can be powered components such as fans (such as P12V_FAN), hard drives, and modules. A load hot-swap event refers to the operation or event of a load being connected to or disconnected from a circuit while the circuit or device is powered on.

[0033] The Baseboard Management Controller (BMC) determines whether a hot-swap event has occurred by identifying the voltage level of the load power circuit's presence detection pin. The BMC communicates directly with the load power circuit interface via the server's internal hardware management bus, monitoring the load's physical connection status in real time. When a load is inserted, the presence detection pin on the load power circuit interface generates a fixed voltage level signal through hardware circuitry (e.g., a high voltage level indicates "in place" and a low voltage level indicates "out of place"). The BMC reads this voltage level signal via the hardware management bus to determine the physical connection status.

[0034] In some embodiments, the baseboard management controller can monitor the hot plug event of the load by at least one of the following implementation methods: connecting a small resistance sampling resistor in series in the load power supply circuit, amplifying the voltage across the resistor by an operational amplifier, and triggering an interrupt signal when the voltage suddenly exceeds a threshold value, thereby determining that a hot plug event of the load has been monitored; or connecting a voltage sampling circuit in parallel to the load power supply end, setting a normal voltage range, and determining that a hot plug event has been detected when the voltage exceeds the range and the rate of change exceeds a threshold value; or determining whether it is a hot plug event by comparing the difference between the real-time current of the load and the rated value, etc. For example, if the fan current drops sharply from the rated value to 0 and the speed signal disappears, it can be determined that the fan is unplugged. If the fan current rises sharply from 0 to the rated value and the speed signal appears, it can be determined that the fan is plugged in. Query instructions can also be sent to the load periodically. If there is no response for multiple consecutive times or the response suddenly resumes, it can be determined that a hot plug event has occurred.

[0035] The above-described embodiment monitors load hot-plug events by capturing subtle signals of sudden current changes through sampling resistors and op amps. This, combined with voltage change rate, load rating comparison, and an active query mechanism, creates a dual-factor detection logic of passive sensing and active verification. This effectively distinguishes parameter changes caused by hot-plugging from fluctuations during normal operation, reducing false positives. Different monitoring methods can be used to match different load types, allowing users to flexibly select or combine them based on load characteristics, eliminating the need to rely on a single monitoring mode. By capturing sudden current changes and voltage anomalies caused by hot-plugging in real time, an interrupt signal or a judgment mechanism can be quickly triggered, allowing sufficient response time for subsequent inrush current suppression. This prevents the instantaneous inrush current from damaging power supply circuit components and prevents voltage surges caused by sudden load removal from affecting other equipment. The active query mechanism complements passive parameter monitoring. Even if passive monitoring experiences blind spots due to varying load characteristics, regular communication queries can accurately detect hot-plug events by detecting response loss / recovery, avoiding the limitations of a single monitoring method and ensuring stable hot-plug event monitoring across diverse load types and power supply environments.

[0036] In some embodiments, when a baseboard management controller detects a load hot-swap event, it monitors the load interface current, load interface temperature, and load interface voltage. It then calculates the current change rate based on the load interface current, the contact temperature rise rate based on the load interface temperature, and the voltage change rate based on the load interface voltage.

[0037] Optionally, a high-frequency current sensor is used to collect the load interface current and calculate the current rate of change (di / dt). A temperature detection module is used to collect the load interface temperature and calculate the contact temperature rise rate (ΔT / Δt). The temperature detection module may include a surface-mount negative temperature coefficient (NTC) thermistor. A voltage detection module is used to collect the load interface voltage and calculate the voltage rate of change (ΔV / Δt). The voltage detection module may include a high-voltage differential probe. Optionally, the load interface current, load interface temperature, and load interface voltage are preprocessed and converted into digital signals through analog-to-digital conversion.

[0038] Specifically, a high-frequency current sensor is connected in series or contactlessly to the load power supply circuit. The sensor bandwidth must match the load current's rate of change, allowing for real-time acquisition of the instantaneous current signal from the load interface. The current signal output by the high-frequency current sensor is conditioned by an op amp and fed into an analog-to-digital converter (ADC) for high-frequency sampling. The digital signal processing unit then performs a differential operation on the continuously sampled current values, outputting a real-time di / dt value, where di is the current difference and dt is the sampling interval.

[0039] A chip NTC thermistor is mounted near the load interface contacts. A voltage divider circuit converts the NTC resistance change into a collectible voltage signal, which is then connected to the signal conditioning circuit. Based on the NTC thermistor's temperature-resistance characteristic curve, the collected resistance signal is first converted to a corresponding temperature value. A sliding window algorithm is then used to calculate the temperature change per unit time, yielding ΔT / Δt. ΔT is the temperature difference, and Δt is the calculation period, typically set to 10 to 100 milliseconds.

[0040] A high-voltage differential probe is connected in parallel across the load interface. The probe must have a high common-mode rejection ratio to resist interference from the power supply circuit. It directly samples the instantaneous voltage across the interface and outputs a differential signal to the subsequent processing unit. The voltage signal output by the high-voltage differential probe is attenuated and filtered before being synchronously sampled by an ADC. The continuous voltage values ​​are processed using a differential operation to calculate ΔV / Δt, where ΔV represents the voltage drop and Δt represents the calculation period. Software algorithms are used to eliminate baseline drift caused by power supply voltage fluctuations.

[0041] The above embodiment, through the combination of a high-frequency current sensor and a high-speed ADC, can capture the current mutation at the moment of hot plugging and improve the response speed. By cross-validating the three types of dynamic parameters, di / dt, ΔT / Δt, and ΔV / Δt, it can effectively eliminate the interference caused by fluctuations in a single parameter.

[0042] In some embodiments, it is first determined whether the current change rate is greater than a first current change rate threshold. When the current change rate is greater than the first current change rate threshold, the load power supply circuit is driven to connect a current limiting resistor to suppress the maximum peak current. limit A / μs, Ilimit = Overcurrent protection value of the power chip / number of loads powered, such as 20 A / μs. I limit This adapts to power supply scenarios with varying load counts, eliminating the need to redesign the current-limit threshold as the number of loads increases or decreases. The current-limiting resistor has a value of 10Ω. Connecting the current-limiting resistor to the load power supply circuit defines pre-charge mode. Pre-charge mode completes the initial charge of the load, preventing cold start shock.

[0043] The fan power supply circuit includes a digitally controlled resistor array and a resistor-capacitor network. The digitally controlled resistor array consists of a digital potentiometer and a MOSFET array; the resistor-capacitor network consists of a capacitor bank and a resistor-capacitor topology. The digitally controlled resistor array can flexibly adjust the equivalent resistance value, providing a stable 10Ω resistance during current limiting. It can also dynamically adjust based on current changes after the load starts. For example, after the load enters steady-state operation, the current limiting resistance value can be reduced or current limiting can be discontinued to avoid power loss and heat generation caused by continuous current limiting. The resistor-capacitor network helps suppress voltage spikes during sudden current changes and buffers circuit impedance changes when the current limiting resistor is connected, minimizing the impact on the load motor's starting performance and preventing load startup failure or speed fluctuations caused by sudden impedance changes.

[0044] When the current change rate is less than or equal to the first preset change rate, it is determined whether the contact temperature rise rate is greater than a preset temperature rise threshold. The preset temperature rise rate may be 10°C / μs. If so, proceed to step S202.

[0045] The above embodiment first uses the first current change rate threshold as the trigger condition, and only intervenes in the regulation when the current suddenly changes and reaches the risk threshold, avoiding unnecessary current limiting during normal operation that affects the load operation; when the trigger condition is met, the load power supply circuit is quickly driven to connect to the 10Ω current limiting resistor, which can directly suppress the maximum peak current at the moment of hot plugging. Combined with the microsecond response of the previous high-frequency current sensor, current limiting can be completed before the surge current reaches the overcurrent protection value of the power supply chip, greatly reducing the probability of the power supply chip triggering protection or damage due to overcurrent. The current limiting resistor is adjusted through a digitally controlled resistor array, which has a faster response than pure software current limiting. Even if the digital signal processing unit suffers a short-term fault, the hardware current limiting mechanism can still work independently, further improving the operational reliability of the load power supply circuit in hot plug scenarios.

[0046] When the current rate of change is less than or equal to a first current rate of change threshold, or when the current rate of change is less than or equal to the first current rate of change threshold and the contact temperature rise rate is less than or equal to a preset temperature rise threshold, or when the contact temperature rise rate is less than or equal to a preset temperature rise threshold, the load power supply circuit switches to steady-state mode. Steady-state mode combines pre-charge mode with a preset RC network in parallel, such as a 5Ω resistor and a 100nF capacitor. The 5Ω resistor stabilizes the loop impedance and suppresses small current fluctuations during steady-state load operation. The 100nF capacitor quickly absorbs voltage spikes, such as those caused by transient grid disturbances and slight load power variations, keeping the supply voltage ripple to a minimum, typically below 10mV. Standardized RC parameters reduce hardware design complexity, eliminating the need to customize RC specifications for different loads. Furthermore, the parallel RC network structure facilitates later parameter adjustments based on load characteristics (for example, increasing the capacitor to 220nF in high-interference environments), providing excellent scalability.

[0047] From a technical perspective, on-demand switching is achieved through multi-dimensional conditional determination, balancing protection and energy consumption. Three types of switching conditions are set: current change rate compliance, current change rate and contact temperature rise rate dual compliance, and contact temperature rise rate compliance. This avoids protection loss caused by premature switching. The system can also flexibly adapt to different load conditions. For example, for low-power loads, only the contact temperature rise rate may meet the standard, while high-power loads may require the current change rate as an auxiliary determination, preventing switching delays caused by over-reliance on a single condition.

[0048] S202 : Input the current change rate, the contact temperature rise rate, and the voltage change rate into a surge probability prediction model to obtain a surge occurrence probability output by the surge probability prediction model.

[0049] The surge probability prediction model is built using a long short-term memory (LSTM) network and trained using a historical dataset. Optionally, the historical dataset includes historical current change rates, contact temperature rise rates, load interface voltages, surge signatures, ambient temperatures, and impedance parameters of the load power supply circuit. The training process for the surge probability prediction model involves first acquiring the historical dataset and using it as training samples to adjust the initial model built using the LSTM network to obtain a converged surge probability prediction model.

[0050] Specifically, the historical dataset is first preprocessed, including but not limited to removing outliers, filling missing values, standardizing, and labeling. For example, using the 3σ principle or boxplot method, extreme data, such as outliers with current change rates far exceeding physical limits and negative impedance values ​​caused by sensor failures, are removed to prevent interference with model training. Short-term missing data is supplemented using linear interpolation, and long-term missing segments are marked as "invalid samples" and removed. Current change rate, heating rate, voltage, ambient temperature, and impedance are normalized using Z-score to transform them into datasets with a mean of 0 and a variance of 1, eliminating the impact of dimensional differences on LSTM model weight updates. "Historical surge tags" are converted into binary labels, with 1 indicating a surge and 0 indicating no surge, ensuring a one-to-one correspondence between labels and the timestamps of the feature data.

[0051] Next, a sequence window is constructed and the preprocessed historical dataset is split. Based on the temporal characteristics of surge events (e.g., hot-plug surges typically occur within 10-100 μs), the sequence window length is set, and a sliding window is used to generate training samples to ensure that the model can predict future surges based on historical time series characteristics. The preprocessed historical dataset is divided into a training set for model parameter updates, a validation set for hyperparameter adjustment, and a test set for final performance evaluation in a 7:2:1 ratio. This division maintains temporal continuity, for example, by taking the first 70% of the dataset as the training set, the middle 20% as the validation set, and the last 10% as the test set to prevent data leakage.

[0052] Then, we build an initial model based on LSTM. The core structure consists of an input layer, an LSTM layer, a fully connected layer, and an output layer. The input dimensions of the input layer include batch size, sequence window length, and number of features. The LSTM layer can be configured with two layers of LSTM units. The first layer enables the return sequence output as a time series sequence, while the second layer disables the return sequence and outputs a single time series feature vector. A dropout layer is also added to prevent overfitting. The fully connected layer maps the feature vector output by the LSTM layer into low-dimensional features. The output layer uses a single output unit with a sigmoid activation function, and the output indicates the probability of a surge occurring after the current time series window ends.

[0053] Further batch training is performed, stopping when the validation set loss converges. During each training round, the model calculates predicted probabilities through forward propagation, compares them with the true labels to calculate the loss, and then uses backpropagation through time (BPTT) to update the weights and biases of the LSTM and fully connected layers to minimize the loss function. Based on the validation set performance, key hyperparameters are adjusted, such as the number of LSTM units, sequence window length, and dropout ratio. The hyperparameter combination that yields the highest F1 score on the validation set is selected. Loss, precision, and recall are monitored in real time on the training and validation sets. If the training set loss continues to decrease while the validation set loss increases, this indicates model overfitting and requires increasing the dropout ratio or reducing the number of LSTM units. Changes in the learning rate are recorded for each training round to ensure that the learning rate remains stable at a low level in the later stages to prevent model oscillation. The trained model is inferred on the test set to calculate precision, recall, and the area under the receiver operating characteristic curve (AUC), which assesses the discriminative ability of the binary classification model. The overall performance of the model is quantified by plotting an ROC curve with the false positive rate (FPR, the proportion of non-surge events mistakenly identified as surges) on the horizontal axis and the true positive rate (TPR, the proportion of surge events correctly identified as surges) on the vertical axis. The area under the curve (AUC) between the curve and the horizontal axis is calculated. An AUC ≥ 0.95 indicates that the model can effectively distinguish between surge and non-surge events.

[0054] The training process of the above-mentioned surge probability prediction model, through full-process optimization of data, model, training and verification, can accurately capture surge events with a low misjudgment rate, providing reliable early warning for surge protection in scenarios such as circuit hot plugging; the full-process monitoring, optimization and verification mechanism ensures stable model training and satisfactory performance, reduces the uncertainty risk in actual applications, and provides a solid predictive foundation for subsequent protection measures.

[0055] In some embodiments, when the current change rate is greater than a first current change rate threshold, the current change rate (di / dt), the contact temperature rise rate (ΔT / Δt), and the voltage change rate (ΔV / Δt) are input into a surge probability prediction model to obtain the surge occurrence probability output by the surge probability prediction model.

[0056] In other embodiments, when the current rate of change is less than or equal to a first current rate of change threshold and the contact temperature rise rate is greater than a preset temperature rise threshold, the current rate of change (di / dt), the contact temperature rise rate (ΔT / Δt), and the voltage rate of change (ΔV / Δt) are input into a surge probability prediction model to obtain a surge probability output by the surge probability prediction model. The contact temperature rise rate can be 10°C / μs, which serves as a trigger condition. This allows for timely intervention when a dangerous contact temperature rise trend occurs, while avoiding frequent model triggering due to a low threshold, such as 5°C / μs, or delayed intervention due to a high threshold, such as 15°C / μs.

[0057] In the above embodiment, if the current rate of change is less than or equal to the first current rate of change threshold and the contact temperature rise rate is greater than the preset temperature rise threshold, the initial prediction is a surge precursor. At this point, the model prediction is triggered, reducing the frequency of model invocation in non-surge scenarios and reducing computing resource consumption. di / dt, ΔT / Δt, and ΔV / Δt are the instantaneous dynamic characteristics preceding a surge. Their combination allows the model to capture both long-term environmental impacts and instantaneous risk mutations. This linkage allows the model to accurately determine surge probability and avoid missed detections.

[0058] After applying the surge probability prediction model to infer the surge probability, the system determines whether the surge probability is greater than or equal to a first preset probability. If so, the load power supply circuit is driven to upgrade from pre-charge mode to steady-state mode. For example, assuming the surge probability is greater than or equal to 90%, a 10Ω resistor is connected to the load power supply circuit as a basic protection element, and a 5Ω resistor is connected in parallel with a 100nF capacitor. With a probability of over 90%, the combination of the 10Ω resistor, the 5Ω resistor, and the 100nF capacitor: the 10Ω resistor handles residual surges before the upgrade, the 5Ω resistor meets the low-loss requirements of steady-state power supply, and the 100nF capacitor ensures power supply stability, further enhancing current limiting and filtering capabilities to prevent surge peak current from breaking down circuit components.

[0059] If the surge probability is less than the first preset probability, the system then determines whether the surge probability is greater than or equal to the second preset probability, and whether the second preset probability is less than the first preset probability. For example, if the first preset probability is 90% and the second preset probability is 80%, a current-limiting resistor is connected to the load power supply circuit. If the surge probability is greater than 80% and less than 90%, a 10Ω current-limiting resistor is connected to prevent excessive loop impedance and power loss caused by excessive power consumption.

[0060] Here, by setting a second preset probability as the medium-risk threshold, the current-limiting resistor is actively connected to the drive circuit when the surge probability is in the medium-risk range of 80%-90%. This preventative measure of early intervention can effectively suppress potential surge currents and prevent them from reaching the trigger threshold of the eFuse, reducing the possibility of eFuse misoperation at the source. Hierarchical regulation based on probability prediction can take targeted measures at the budding stage of risks, avoiding overprotection in low-risk situations and reducing the surge intensity in high-risk scenarios through the early connection of current-limiting resistors. This makes the protection response more accurate and timely, and compensates for the delayed response of the eFuse thermal shutdown function.

[0061] S203 : When the surge occurrence probability is greater than or equal to a first preset probability, calculate resistance and capacitance parameters of the load power supply circuit according to the monitored load interface current, load interface voltage, and load interface temperature.

[0062] This application monitors the current change rate, contact temperature rise rate, and voltage change rate in real time, and uses a surge probability prediction model based on a long short-term memory network to determine the probability of a surge in advance. It intervenes in advance when the probability of a surge is greater than or equal to a first preset probability, thus avoiding false triggering caused by the passive response of the eFuse. When a high surge risk is predicted, the resistance and capacitance parameters of the load power supply circuit are calculated in real time and actively adjusted, so that the circuit state can be adjusted before a surge occurs. Compared with the passive mechanism of the eFuse that relies on its own temperature rise to trigger protection, the response speed is greatly improved, and it can respond to instantaneous changes in fast hot-swap scenarios in a timely manner, avoiding equipment damage caused by untimely protection.

[0063] In some embodiments, when the probability of a surge occurring is greater than or equal to a first preset probability, a spectrum analysis is first performed on the load interface current to extract the surge main frequency f; then, the target impedance is calculated based on the load interface voltage and the limiting current, and then, based on the surge main frequency f and the target impedance, the capacitance value of the load power supply circuit is calculated; and then, based on the temperature compensation coefficient corresponding to the load interface voltage, the target current, and the load interface temperature, the resistance value of the load power supply circuit is calculated.

[0064] Among them, the surge main frequency is the frequency component with the strongest energy in the surge current. target =V / I limit , V is the load interface voltage; I limit =Overcurrent protection OCP value of the power supply chip ÷ number of loads powered, indicating the maximum current allowed by the load power supply circuit.

[0065] Specifically, the PID control chip performs fast Fourier transform to analyze the current spectrum, finds the surge main frequency f, and then calculates the target impedance based on the load interface voltage and the limit current, and then calculates the target impedance Z targetand surge frequency f, calculate the capacitance value C of the load power supply circuit according to the following formula (1): opt : (1) According to the following formula (2), based on the load interface voltage V, target current I target Temperature compensation coefficient K corresponding to the load interface temperature temp , calculate the resistance value R of the load power supply circuit opt : (2) The target current I in formula (2) target It is the rated current when the load power supply circuit is working normally. Temperature compensation coefficient K temp Used to correct the effect of temperature on resistance and ensure that the calculated resistance value can meet the requirements at different temperatures.

[0066] In the above embodiment, when the surge probability is ≥ a first preset probability, the PID control chip first analyzes the current spectrum and extracts the surge frequency f through a fast Fourier transform (FFT) analysis. This avoids capacitor and resistor matching errors caused by using fixed frequency parameters. The PID control chip ensures that the RC parameter calculations are completed before the surge energy erupts, allowing sufficient time for subsequent RC network adjustments and preventing protection delays caused by lengthy analysis. Through spectrum analysis, FFT extraction, and RC calculation, the RC parameters are dynamically adjusted based on surge characteristics, load conditions, and temperature, improving surge suppression effectiveness. This customized parameter control based on surge characteristics can specifically suppress the frequency band where surge current energy is concentrated, reducing the surge intensity at its source and preventing it from reaching the eFuse trigger threshold, thereby minimizing the possibility of false tripping. The entire control process is implemented through real-time monitoring of current, voltage, and temperature, FFT, and precise calculation of RC parameters, ultimately leading to dynamic adjustment. This significantly improves response speed compared to the passive triggering mode of the eFuse, which relies on its own temperature rise. In particular, the introduction of target impedance and temperature compensation coefficient ensures that the power supply circuit parameters are always in the optimal protection state during the instantaneous changes of fast hot plugging, solving the problem of untimely protection caused by delayed response of eFuse.

[0067] In some embodiments, when the probability of surge occurrence is greater than or equal to a first preset probability, the load interface current is first subjected to spectrum analysis to extract the surge main frequency f; then the target impedance is calculated based on the load interface voltage and the limiting current, and then the target impedance Z is calculated based on the target impedance Z. target and surge frequency f, the solution satisfies The resistance and capacitance parameters. Because the resistor consumes power (P=I 2 R), under the premise of meeting the impedance requirements, the resistance and capacitance parameters with smaller resistance can be selected to control the load power supply circuit.

[0068] The above embodiment extracts the surge main frequency f through spectrum analysis and calculates the resistor and capacitor parameters that meet the target impedance based on the impedance formula. This can accurately match the frequency characteristics of the surge and ensure that the total impedance Z after the resistor and capacitor are connected in series is equal to the target impedance, avoiding insufficient or excessive surge suppression due to total impedance deviation. This customized impedance design can effectively attenuate the energy of the surge current and prevent it from reaching the trigger threshold of the eFuse, thereby reducing false operations from the root. At the same time, selecting a smaller resistor while meeting the impedance requirements not only ensures the surge suppression effect, but also avoids normal operation interference caused by excessive current limiting, resolving the contradiction between eFuse protection and normal equipment operation. Compared with the passive triggering mechanism of the eFuse that relies on its own temperature rise, this forward-looking regulation based on surge characteristics has a faster response speed and can complete the circuit adjustment before the surge energy dissipates, making up for the protection lag defect of the eFuse caused by thermal conduction delay. It is particularly suitable for fast transient scenarios when hot-plugging fans.

[0069] S204: Regulate the load power supply circuit according to resistance and capacitance parameters.

[0070] The RC parameters include resistance and capacitance values. The load power supply circuit includes a digitally controlled variable resistor array and a RC network. The digitally controlled variable resistor array consists of a digital potentiometer and a power MOSFET array, capable of rapidly switching resistance values ​​to limit current amplitude. The RC network consists of a capacitor bank and a RC topology. The capacitor bank supports switching capacitance via a control circuit. The RC topology refers to the connection method of resistors and capacitors. The RC network can dynamically match RC parameters, suppress inrush current, and reduce unstable oscillations caused by high-frequency signals in the circuit.

[0071] The digitally controlled resistor array of the load power supply circuit is regulated according to the resistance value; the resistance and capacitance network of the load power supply circuit is regulated according to the capacitance value. Specifically, when the impedance formula is solved to calculate the target resistance value R opt Afterwards, the main control unit sends an 8-bit control command to the digitally controlled resistor array via the SPI / I2C communication protocol. The first 4 bits control the MOSFET switch state, and the last 4 bits control the digital potentiometer resistance value. After receiving the command, the MOSFET switch completes the on / off process within 100ns, and the digital potentiometer synchronously adjusts the sliding end position to achieve a combination of coarse adjustment of the fixed resistor combination and fine adjustment of the digital potentiometer to achieve the target resistance value. The final resistance error is controlled within ±0.01Ω. According to the calculated C opt The real-time main frequency f is matched with the control signal sent through the GPIO port, driving the analog switch to complete the on / off control within 50ns, thereby combining the target capacitance value. At the same time, according to the voltage-capacitance characteristic of the varactor diode, its reverse bias voltage is adjusted to achieve pF-level fine-tuning, and the final capacitance error is controlled within ±0.1nF.

[0072] For example, in the scenario of f=2MHz and Ztarget=6Ω, the precise combination of R=5Ω and C=13nF can reduce the peak surge current from 30A to 2A (I=V / Z=12V / 6Ω=2A), which is completely within the OCP value of the power chip (such as 4A) and avoids triggering overcurrent protection.

[0073] In any of the above embodiments, the system bus voltage is monitored and a voltage difference between the system bus voltage and the load interface voltage is calculated. If the voltage difference is less than or equal to a voltage difference threshold and the current rate of change is less than a second current rate of change threshold for a predetermined period of time, the resistance of the load power supply circuit in the steady-state mode is reduced.

[0074] Among them, the system bus voltage V bus It is the energy source for powering the load. The load interface voltage V fan The voltage difference threshold can be set to 0.8V, the second current change rate threshold can be set to 1Aμs, and the preset time length can be set to 10μs.

[0075] Specifically, the system bus voltage V bus , calculate the system bus voltage V bus Interface voltage with load V fan The voltage difference ΔV between them. When ΔV≤0.8V and the current change rate di / dt<1A / μs within 10μs, it indicates that the voltage drop from the bus to the load is small, and the resistance value of the load power supply circuit is reduced from 5Ω to 1Ω. The bus has sufficient power supply capacity. At this time, reducing the resistance will not cause V fan Too low.

[0076] The resistance reduction trigger condition is that the voltage difference is less than or equal to the voltage difference threshold and the current change rate within a preset time period is less than a second current change rate threshold. This ensures that resistance adjustment is only performed in scenarios with stable power supply and no surge risk. This avoids circuit safety hazards caused by misoperation, reduces power loss, improves power supply efficiency, and ensures load operation stability.

[0077] If the voltage difference is greater than the voltage difference threshold, or the current change rate is less than the second current change rate threshold and the duration of this situation is less than the preset duration, or the current change rate is greater than the second current change rate threshold, then a delay of 50μs is made to maintain the steady-state mode of the load power supply circuit, and the voltage difference and current change rate are re-judged after 50μs.

[0078] If ΔV > 0.8V, the bus power supply capacity is insufficient. Blindly reducing the resistance will further increase the current (I = Vfan / R), causing ΔV to continue to expand and even triggering bus overcurrent protection. Therefore, a 50μs delay is applied to maintain the load power supply circuit in steady-state mode. Assuming the current rate of change remains stable at 0.5A / μs for 10μs, this indicates no transient surges and the load power supply circuit can maintain steady-state mode. If the current rate of change briefly drops below 1A / μs but does not last for 10μs, this may be current adjustment before a surge. Not adjusting the resistance at this time can prevent current runaway caused by insufficient resistance during a surge.

[0079] In some embodiments, the remaining surge energy in the load power supply circuit is monitored. If the remaining surge energy exceeds a preset energy threshold, the remaining surge energy is stored in a storage capacitor bank via an energy feedback circuit. Optionally, the remaining current i is collected by a high-frequency current sensor and combined with the loop resistance R to calculate the remaining surge energy in real time.

[0080] The residual surge energy is the instantaneous energy remaining after surge suppression. The preset energy threshold can be 30%. The 30% setting ensures that the recovered energy has practical utilization value while avoiding the situation where the startup cost of the feedback circuit is higher than the recovery benefit due to the residual energy being too low.

[0081] The energy feedback circuit can be an H-bridge circuit, such as Figure 3 As shown, when the remaining surge energy is greater than the preset energy threshold, control paths S1 and S4 are turned on, so that the remaining surge energy is transferred from the load power supply circuit to the energy storage capacitor group. The energy storage capacitor group can be a supercapacitor. Among them, a supercapacitor is a new type of energy storage device between traditional capacitors and batteries. Its core features are that its specific capacity is much higher than that of ordinary capacitors and its charging and discharging speed is much faster than that of batteries. Its energy storage principle is based on the double-layer capacitance at the interface between the electrode and the electrolyte or the Faraday pseudo-capacitance on the electrode surface. It can achieve millisecond-level rapid charging and discharging, a cycle life of tens of thousands to hundreds of thousands of times, and a wide operating temperature range.

[0082] Residual surge energy monitoring can directly reflect the energy status. This closed loop can avoid the secondary impact of residual surge energy on the circuit and improve the utilization rate of surge energy.

[0083] If the remaining surge energy is less than or equal to the preset energy threshold, the load power supply circuit is regulated to steady-state mode. Remaining surge energy ≤ 30% means that the residual surge energy has fallen to a safe range, and the circuit is unlikely to experience a secondary energy shock.

[0084] For example, if the total surge energy is 100mJ and the remaining 25mJ is ≤30mJ, this energy can be discharged through a 5Ω resistor, resulting in only a brief 0.3°C temperature rise and no damage to the load. During the surge protection phase, a high-resistance configuration is used to suppress peak energy. While this effectively limits current, it also results in high power loss. A low-resistance configuration, when controlled in steady-state mode, reduces losses. Combined with capacitor filtering to reduce voltage fluctuations, this configuration reduces overall circuit energy consumption after the surge.

[0085] In the above embodiment, when the remaining energy is greater than 30%, the H-bridge circuit is used to recover the energy to the supercapacitor group first to avoid energy waste. When the energy is less than or equal to 30% and the recovery value is low, the system directly switches to the steady-state mode to avoid the startup cost of the feedback circuit. The two modes cover the full range of scenarios of high energy recovery and low energy steady state, thereby improving the energy utilization rate after the surge.

[0086] After the energy storage capacitor group stores a certain amount of surge energy, when it is monitored that the system bus voltage is less than the voltage threshold, or the load operating parameters are less than the required parameters, the energy in the energy storage capacitor group is transferred to the load power supply circuit through the energy feedback circuit. For example, when the system bus voltage is less than 11V, or the fan speed is less than the preset speed, the energy feedback circuit conduction paths S2 and S3 are controlled, and the energy storage capacitor group releases energy. When the system bus voltage drops to 11V briefly, the load power supply may be insufficient, and the energy storage capacitor group releases energy to maintain the load voltage stable, avoiding the protection action of triggering the resistor adjustment. When the fan speed is less than the preset speed, energy feedback can directly supplement the load current to push the speed back to the rated value, avoiding the decrease in heat dissipation efficiency due to insufficient speed, and ensuring the normal function of the load.

[0087] When the load interface voltage is detected as a back EMF, the current direction of the energy feedback circuit is controlled to suppress the back EMF. For example, when a fan is unplugged, the power supply circuit generates a -15V back EMF. The energy feedback circuit conducts paths S2 and S4, directing the energy corresponding to the back EMF into the energy storage capacitor bank, suppressing the voltage spike. This achieves active suppression and reuse of back EMF.

[0088] When an inductive load like a fan is unplugged, the inductive component generates a -15V back EMF due to the sudden current drop. If left untreated, this could potentially breakdown the MOSFET and power supply chip in the power supply circuit. However, the conduction of paths S2 and S4 establishes a reverse energy path from the load power supply circuit to the energy feedback circuit and then to the energy storage capacitor bank within 1-2μs. This rapidly transfers the energy corresponding to the -15V back EMF to the energy storage capacitor bank, suppressing voltage spikes and improving energy recovery.

[0089] Based on any of the above embodiments, after the hot plug event is completed, the load power supply circuit is adjusted to a low power mode. Exemplarily, in the low power mode, the resistance value R=0.1Ω and the capacitance value C=470μF of the load power supply circuit. After the hot plug is completed, the circuit enters a long-term steady state and is adjusted to a low power mode, thereby reducing the long-term operating cost of the equipment, while avoiding the circuit temperature rise caused by high resistance heating and extending the life of the components. Although there is no sudden surge after the hot plug is completed, there may be slight fluctuations in the power grid and slight power changes in the load. The ultra-large capacitance of the 470μF capacitor can provide extremely strong energy buffering capacity to avoid performance deviations of the load due to power supply fluctuations.

[0090] In some embodiments, parameters in the process of regulating the load power supply circuit, including but not limited to impedance parameters and residual surge energy, are uploaded to a database for use in updating the surge probability prediction model.

[0091] In summary, the control method for the load power supply circuit provided in the embodiment of the present application comprehensively reflects the rapid change of the circuit state at the moment of hot plugging by obtaining the three key indicators of the current change rate, contact heating rate and voltage change rate of the load interface when a hot plug event occurs. Subsequently, these change rates are input into the surge probability prediction model to obtain the probability of surge occurrence. Unlike the traditional method of relying solely on eFuse to passively deal with overcurrent and overvoltage, this method can predict surge risks in advance. Because the instantaneous surge current generated during hot plugging is the key factor causing the eFuse to malfunction, by predicting the surge probability in advance, when the surge is likely to occur, measures can be taken to prevent the eFuse from malfunctioning due to surge impact, thereby protecting the circuit from damage.

[0092] When the surge probability is greater than or equal to the first preset probability, it means that there may be a greater surge risk. At this time, the method calculates the resistance and capacitance parameters of the load power supply circuit based on the monitored load interface current, load interface voltage, and load interface temperature. When the power supply fluctuates greatly or the design is unstable, the eFuse overvoltage protection capability is insufficient, and by calculating the resistance and capacitance parameters in real time, the impedance characteristics of the circuit can be flexibly adjusted. For example, when the power supply voltage rises, by changing the capacitance or resistance value, the circuit current is adjusted to share part of the overvoltage energy, thereby compensating for the defects of the eFuse overvoltage protection in a complex power supply environment and providing more comprehensive overvoltage protection for the load power supply circuit.

[0093] The key execution step of this method is to regulate the load power supply circuit based on the calculated resistance and capacitance parameters. Unlike the eFuse thermal shutdown function, which relies on the internal temperature rise of the IC and has a slow response speed, this regulation is real-time and based on dynamic calculation of multiple parameters. For example, when a rapid hot plug or short circuit event occurs, the circuit can be quickly adjusted according to the calculated resistance and capacitance parameters. For example, by changing the connection method of the capacitor or resistor through the switching element, the circuit characteristics can be quickly changed, and abnormal changes in current and voltage can be suppressed in time. This greatly improves the protection response speed, effectively avoids equipment damage caused by untimely protection, and ensures the stable operation of the server in complex situations such as load hot plug.

[0094] like Figure 4 As shown, Figure 4 The flowchart of the dynamic switching of the resistance and capacitance parameters of the load power supply circuit provided in the embodiment of the present application first obtains the current change rate, contact heating rate and voltage change rate of the load interface, and determines whether the current change rate of the load interface is greater than or equal to the first current change rate threshold I limit If so, it indicates that the load power supply circuit meets the pre-charge trigger condition. A 10Ω current-limiting resistor is first connected, which is the initial current limiting process. The current change rate, contact temperature rise rate, and voltage change rate of the load interface are then input into the surge probability prediction model to obtain the surge occurrence probability. A determination is made as to whether the surge occurrence probability is greater than or equal to a second preset probability of 80%. If so, it indicates that the load power supply circuit meets the pre-charge trigger condition, and a 10Ω current-limiting resistor is connected. Furthermore, a determination is made as to whether the surge occurrence probability is greater than or equal to a first preset probability of 90%. If so, a 5Ω resistor is connected in parallel to the load power supply circuit, and a 100nF capacitor is connected. Next, a determination is made as to whether the steady-state switching condition is met: whether the voltage difference between the system bus voltage and the load interface voltage is less than or equal to a voltage difference threshold of 0.8V, and whether the current change rate within a preset duration of 10μs is less than a second current change rate threshold of 1A / μs. If the voltage difference is ≤0.8V and less than 1A / μs within 10μs, the steady-state switching condition is met, and the load power supply circuit is switched to the 1Ω resistor. If the steady-state switching conditions are not met, the 5Ω resistor and 100nF capacitor configuration is maintained, and after a 50μs delay, the steady-state switching conditions are re-evaluated. This process adjusts the resistor and capacitor parameters in stages, combining indicators such as the current rate of change, power prediction, and voltage rate of change to ensure a smooth transition from pre-charge to steady-state operation, preventing sudden current and voltage changes from impacting circuit components while optimizing energy efficiency and stability.

[0095] like Figure 5 As shown, Figure 5 Schematic diagram of the process of controlling the load power supply circuit provided in the embodiment of the present application Figure 2. When a hot plug event occurs, real-time signals such as current, temperature, and voltage of the load power supply circuit are collected through high-frequency current sensors, temperature sensors, and voltage sensors. The collected real-time signals are preprocessed, and then the current change rate di / dt, contact temperature rise rate, and voltage change rate are calculated and the spectrum is analyzed. Subsequently, it is determined whether the current change rate is greater than the first current change rate threshold of 20A / μs. If so, the current change rate, contact temperature rise rate, and voltage change rate are input into the surge probability prediction model to obtain the probability of surge occurrence. If not, it is further determined whether the contact temperature rise rate is greater than the preset temperature rise threshold of 10℃ / μs. When the contact temperature rise rate is greater than the preset temperature rise threshold, the current change rate, contact temperature rise rate, and voltage change rate are input into the surge probability prediction model to obtain the probability of surge occurrence. When the contact temperature rise rate is less than the preset temperature rise threshold, the steady-state mode is maintained (5Ω resistor and 100nF capacitor in parallel).

[0096] When the probability of a surge is greater than or equal to 90%, the PID control chip calculates the resistance and capacitance parameters R and C of the load power supply circuit based on the monitored load interface current, load interface voltage, and load interface temperature. The digitally controlled resistor array and resistor-capacitor network of the load power supply circuit are regulated based on the resistance and capacitance parameters.

[0097] The remaining surge energy is then monitored to determine whether it exceeds the energy threshold. If so, the remaining surge energy is recovered and stored in the supercapacitor; if not, the load power supply circuit is regulated to a steady-state mode. Through the above description of the implementation method, those skilled in the art will clearly understand that the method according to the above embodiment can be implemented using software plus the necessary general hardware platform. Of course, it can also be implemented through hardware, but in many cases the former is a more preferred implementation method.

[0098] like Figure 6 As shown, an embodiment of the present application further provides a control device for a load power supply circuit, the device comprising: An acquisition module 601 is configured to acquire a current change rate, a contact temperature rise rate, and a voltage change rate of a load interface when the baseboard management controller monitors a hot plug event of the load. The hot plug event is identified by a level signal from a presence detection pin of a load power supply circuit. Surge prediction module 602, configured to input the current change rate, contact temperature rise rate, and voltage change rate into a surge probability prediction model to obtain a surge occurrence probability output by the surge probability prediction model; wherein the surge probability prediction model is constructed based on a long short-term memory network and trained based on a historical data set; A calculation module 603 is configured to calculate resistance and capacitance parameters of a load power supply circuit based on the monitored load interface current, load interface voltage, and load interface temperature when the surge occurrence probability is greater than or equal to a first preset probability; The control module 604 is used to control the load power supply circuit according to the resistance and capacitance parameters.

[0099] As an optional implementation provided in an embodiment of the present application, the acquisition module 601 is specifically used to: monitor the load interface current, load interface temperature and load interface voltage when the baseboard management controller monitors the hot plug event of the load; calculate the current change rate based on the load interface current, calculate the contact heating rate based on the load interface temperature, and calculate the voltage change rate based on the load interface voltage.

[0100] As an optional implementation provided in an embodiment of the present application, the acquisition module 601 is also used to: determine whether the current change rate is greater than a first current change rate threshold; the control module 604 is also used to: drive the load power supply circuit to connect to the current limiting resistor when the current change rate is greater than the first current change rate threshold.

[0101] As an optional implementation provided in an embodiment of the present application, the surge prediction module 602 is also used to obtain a historical data set, which includes a historical current change rate, a historical contact heating rate, a historical load interface voltage, a historical surge mark, a historical ambient temperature, and a historical impedance parameter of the load power supply circuit; and adjust the initial model constructed based on the long short-term memory network according to the historical data set to obtain a surge probability prediction model.

[0102] As an optional implementation provided in an embodiment of the present application, the surge prediction module 602 is also used to determine whether the current change rate is greater than a first current change rate threshold; when the current change rate is less than or equal to the first current change rate threshold, determine whether the contact heating rate is greater than a preset heating threshold, so that when the contact heating rate is greater than the preset heating threshold, the current change rate, the contact heating rate and the voltage change rate are input into the surge probability prediction model.

[0103] As an optional implementation provided in an embodiment of the present application, the surge prediction module 602 is also used to drive the load power supply circuit to upgrade from the pre-charging mode to the steady-state mode when the probability of a surge occurring is greater than or equal to a first preset probability; wherein the steady-state mode is a preset resistor-capacitor network connected in parallel on the basis of the pre-charging mode.

[0104] As an optional implementation provided in an embodiment of the present application, the control module 604 is also used to: switch the load power supply circuit to a steady-state mode when the current change rate is less than or equal to a first current change rate threshold, and / or the contact heating rate is less than or equal to a preset heating threshold.

[0105] As an optional implementation provided in an embodiment of the present application, the control module 604 is also used to: when the probability of a surge occurring is greater than or equal to a second preset probability and less than the first preset probability, drive the load power supply circuit to connect to a current limiting resistor; wherein the second preset probability is less than the first preset probability.

[0106] As an optional implementation provided in an embodiment of the present application, the calculation module 603 is specifically used to: when the probability of a surge occurring is greater than or equal to a first preset probability, perform spectrum analysis on the load interface current to extract the surge main frequency; calculate the target impedance based on the load interface voltage and the limiting current; calculate the capacitance value of the load power supply circuit based on the surge main frequency and the target impedance; and calculate the resistance value of the load power supply circuit based on the temperature compensation coefficient corresponding to the load interface voltage, the target current and the load interface temperature.

[0107] As an optional implementation provided in an embodiment of the present application, the control module 604 is specifically used to: control the digitally controlled resistor array of the load power supply circuit according to the resistance value; and control the resistor-capacitor network of the load power supply circuit according to the capacitance value.

[0108] As an optional implementation provided in an embodiment of the present application, the control module 604 is also used to: monitor the system bus voltage; calculate the voltage difference between the system bus voltage and the load interface voltage; if the voltage difference is less than or equal to the voltage difference threshold, and the current change rate is less than the second current change rate threshold for a preset time period, then reduce the resistance value of the load power supply circuit in the steady-state mode.

[0109] As an optional implementation provided in an embodiment of the present application, the device also includes an energy feedback module for: monitoring the residual surge energy of the load power supply circuit; when the residual surge energy is greater than a preset energy threshold, storing the residual surge energy in the energy storage capacitor group through the energy feedback circuit.

[0110] As an optional implementation provided in an embodiment of the present application, the control module 604 is further configured to: when the remaining surge energy is less than or equal to a preset energy threshold, control the load power supply circuit to a steady-state mode.

[0111] As an optional implementation provided in an embodiment of the present application, the energy feedback module is also used to: when it is monitored that the system bus voltage is less than the voltage threshold, or the load operating parameters are less than the required parameters, transfer the energy in the energy storage capacitor group to the load power supply circuit through the energy feedback circuit.

[0112] As an optional implementation provided in an embodiment of the present application, the energy feedback module is further used to: when it is detected that the load interface voltage is a reverse electromotive force, suppress the reverse electromotive force by controlling the current direction of the energy feedback circuit.

[0113] As an optional implementation provided in an embodiment of the present application, the control module 604 is further configured to: after the hot plug event processing is completed, control the load power supply circuit to a low power consumption mode.

[0114] For the description of the features in the embodiment corresponding to the control device of the load power supply circuit, reference can be made to the relevant description of the embodiment corresponding to the control method of the load power supply circuit, which will not be repeated here.

[0115] like Figure 7 As shown, an embodiment of the present application also provides an electronic device, including a memory 701 and a processor 702, wherein the memory 701 stores a computer program, and the processor 702 is configured to run the computer program to execute the steps in any of the above-mentioned load power supply circuit control method embodiments.

[0116] An embodiment of the present application further provides a computer-readable storage medium, in which a computer program is stored, wherein the computer program is configured to execute the steps of any of the above-mentioned load power supply circuit control method embodiments when running.

[0117] In an exemplary embodiment, the computer-readable storage medium may include, but is not limited to, various media that can store computer programs, such as a USB flash drive, a read-only memory (ROM), a random access memory (RAM), a mobile hard disk, a magnetic disk, or an optical disk.

[0118] An embodiment of the present application further provides a computer program product, which includes a computer program. When the computer program is executed by a processor, the steps in any of the above-mentioned load power supply circuit control method embodiments are implemented.

[0119] An embodiment of the present application also provides another computer program product, including a non-volatile computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, it implements the steps in any of the above-mentioned load power supply circuit control method embodiments.

[0120] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the above description has generally described the components and steps of each example according to their functions. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0121] The above is a detailed introduction to the control method, device, equipment, medium and program product of a load power supply circuit provided by the present application. Specific examples are used herein to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only applicable to help understand the method of the present application and its core idea. It should be pointed out that for ordinary technicians in this technical field, without departing from the principles of the present application, several improvements and modifications can be made to the present application, and these improvements and modifications also fall within the scope of protection of the claims of the present application.

Claims

1. A method for controlling a load power supply circuit, characterized in that: include: When the baseboard management controller monitors a hot plug event of the load, the current change rate, contact temperature rise rate, and voltage change rate of the load interface are obtained; wherein the hot plug event is identified by a level signal of a presence detection pin of the load power supply circuit; Inputting the current change rate, the contact temperature rise rate, and the voltage change rate into a surge probability prediction model to obtain a surge occurrence probability output by the surge probability prediction model; wherein the surge probability prediction model is constructed based on a long short-term memory network and is trained based on a historical data set; When the surge occurrence probability is greater than or equal to a first preset probability, calculating resistance and capacitance parameters of a load power supply circuit according to the monitored load interface current, load interface voltage, and load interface temperature; The load power supply circuit is regulated according to the resistance and capacitance parameters.

2. The method according to claim 1, characterized in that When the baseboard management controller monitors the hot plug event of the load, obtaining the current change rate, contact temperature rise rate and voltage change rate of the load interface includes: When the baseboard management controller detects a hot plug event of the load, it monitors the load interface current, load interface temperature and load interface voltage; The current change rate is calculated according to the load interface current, the contact temperature rise rate is calculated according to the load interface temperature, and the voltage change rate is calculated according to the load interface voltage.

3. The method according to claim 1, characterized in that After obtaining the current change rate, contact temperature rise rate, and voltage change rate of the load interface, the method further includes: Determining whether the current change rate is greater than a first current change rate threshold; When the current change rate is greater than a first current change rate threshold, the load power supply circuit is driven to connect to a current limiting resistor.

4. The method according to claim 1, wherein When the baseboard management controller monitors the hot plug event of the load, before obtaining the current change rate, contact temperature rise rate and voltage change rate of the load interface, the method further includes: Acquiring the historical data set, wherein the historical data set includes a historical current change rate, a historical contact temperature rise rate, a historical load interface voltage, a historical surge mark, a historical ambient temperature, and a historical impedance parameter of a load power supply circuit; The initial model constructed based on the long short-term memory network is adjusted according to the historical data set to obtain the surge probability prediction model.

5. The method according to claim 1, wherein After acquiring the current change rate, contact temperature rise rate, and voltage change rate of the load interface when the baseboard management controller monitors a hot plug event of the load, and before inputting the current change rate, contact temperature rise rate, and voltage change rate into a surge probability prediction model to obtain a surge occurrence probability, the method further includes: Determining whether the current change rate is greater than a first current change rate threshold; When the current change rate is less than or equal to a first current change rate threshold, determine whether the contact temperature rise rate is greater than a preset temperature rise threshold, and input the current change rate, the contact temperature rise rate and the voltage change rate into the surge probability prediction model when the contact temperature rise rate is greater than the preset temperature rise threshold.

6. The method according to claim 1, characterized in that After inputting the current change rate, the contact temperature rise rate, and the voltage change rate into a surge probability prediction model to obtain a surge occurrence probability output by the surge probability prediction model, the method further includes: When the surge occurrence probability is greater than or equal to the first preset probability, driving the load power supply circuit to upgrade from the pre-charging mode to the steady-state mode; The steady-state mode is to connect a preset resistor-capacitor network in parallel on the basis of the pre-charging mode.

7. The method according to claim 5, characterized in that The method further comprises: When the current change rate is less than or equal to the first current change rate threshold, and / or the contact temperature rise rate is less than or equal to the preset temperature rise threshold, the load power supply circuit is switched to a steady-state mode.

8. The method according to claim 1, characterized in that The method further comprises: When the surge occurrence probability is greater than or equal to a second preset probability and less than the first preset probability, driving the load power supply circuit to connect to a current limiting resistor; The second preset probability is smaller than the first preset probability.

9. The method according to claim 1, characterized in that When the surge occurrence probability is greater than or equal to a first preset probability, calculating resistance and capacitance parameters of a load power supply circuit according to the monitored load interface current, load interface voltage, and load interface temperature includes: When the surge occurrence probability is greater than or equal to the first preset probability, performing spectrum analysis on the load interface current to extract the surge main frequency; Calculating target impedance according to the load interface voltage and the limiting current; Calculating a capacitance value of the load power supply circuit according to the surge main frequency and the target impedance; The resistance value of the load power supply loop is calculated according to the load interface voltage, the target current, and a temperature compensation coefficient corresponding to the load interface temperature.

10. The method according to claim 9, characterized in that The regulating the load power supply circuit according to the resistance and capacitance parameters includes: According to the resistance value, regulating the digitally controlled resistor array of the load power supply circuit; The resistance-capacitance network of the load power supply circuit is regulated according to the capacitance value.

11. The method according to claim 1, wherein The method further comprises: Monitor system bus voltage; Calculating a voltage difference between a system bus voltage and the load interface voltage; If the voltage difference is less than or equal to the voltage difference threshold, and the current change rate is less than the second current change rate threshold for a predetermined period of time, the resistance value of the load power supply circuit in the steady-state mode is reduced.

12. The method according to claim 1, characterized in that After regulating the load power supply circuit according to the resistance-capacitance parameters, the method further includes: monitoring the residual surge energy of the load power supply circuit; When the remaining surge energy is greater than a preset energy threshold, the remaining surge energy is stored in the energy storage capacitor group through the energy feedback circuit.

13. The method according to claim 12, characterized in that The method further comprises: When the remaining surge energy is less than or equal to the preset energy threshold, the load power supply circuit is regulated to a steady-state mode.

14. The method according to claim 12, characterized in that The method further comprises: When it is monitored that the system bus voltage is less than the voltage threshold, or the load operating parameter is less than the required parameter, the energy in the energy storage capacitor group is transferred to the load power supply circuit through the energy feedback circuit.

15. The method according to claim 12, characterized in that The method further comprises: When it is detected that the load interface voltage is a reverse electromotive force, the reverse electromotive force is suppressed by controlling the current direction of the energy feedback circuit.

16. The method according to claim 1, wherein The method further comprises: After the hot plug event is processed, the load power supply circuit is adjusted to a low power consumption mode.

17. A control device for a load power supply circuit, characterized in that: include: When the baseboard management controller monitors a hot plug event of the load, the current change rate, contact temperature rise rate, and voltage change rate of the load interface are obtained; wherein the hot plug event is identified by a level signal of a presence detection pin of the load power supply circuit; Inputting the current change rate, the contact temperature rise rate, and the voltage change rate into a surge probability prediction model to obtain a surge occurrence probability output by the surge probability prediction model; wherein the surge probability prediction model is constructed based on a long short-term memory network and is trained based on a historical data set; When the surge occurrence probability is greater than or equal to a first preset probability, calculating resistance and capacitance parameters of a load power supply circuit according to the monitored load interface current, load interface voltage, and load interface temperature; The load power supply circuit is regulated according to the resistance and capacitance parameters.

18. An electronic device, characterized in that: include: memory for storing computer programs; A processor, configured to implement the steps of the method for controlling a load power supply circuit as claimed in any one of claims 1 to 16 when executing the computer program.

19. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, wherein when the computer program is executed by a processor, the steps of the method for controlling a load power supply circuit according to any one of claims 1 to 16 are implemented.

20. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method for controlling a load power supply circuit as claimed in any one of claims 1 to 16 are implemented.

Images