Adaptive temperature control method for multi-stage power topology and voltage conversion system

Abstract

The present application belongs to the technical field of voltage transformation, and particularly relates to a self-adaptive temperature control method for a multi-stage power topology and a voltage transformation system. First temperature data is generated by collecting real-time temperature of a target device, and a power topology level where a current load is located is divided according to the first temperature data. Then, corresponding voltage adjustment instructions are generated based on the level, and control signals matching input characteristics are generated to drive topology switching. After voltage adjustment, second temperature data is continuously collected, a temperature compensation value is calculated by comparative analysis, and the temperature compensation value is used to update the voltage instructions to form a closed-loop control. Dynamic linkage control between voltage output and temperature change is realized, the response capability of the system to load fluctuation and the thermal management precision are improved, the system instability problem caused by untimely voltage regulation or temperature control lag in the prior art is effectively solved, and thus the overall operation efficiency and reliability are improved.

Description

Technical Field

[0001] This invention belongs to the field of transformer technology, specifically relating to an adaptive temperature control method and transformer system for multi-stage power topologies. Background Technology

[0002] In power electronic systems, especially in multi-level power topologies, changes in load conditions often cause fluctuations in voltage output and temperature characteristics. To maintain stable system operation, existing technologies generally employ open-loop or fixed-level voltage control strategies, combined with independent temperature monitoring modules for over-temperature protection. For example, traditional solutions typically use fixed voltage output levels to handle different load conditions, while relying on temperature sensors to collect device temperatures. When the detected temperature exceeds a safe threshold, passive thermal protection measures such as frequency reduction, current limiting, or shutting down certain circuits are triggered.

[0003] However, this control method has significant limitations: due to the dynamic coupling between voltage regulation and temperature changes, fixed voltage control struggles to adapt to temperature fluctuations caused by rapid load changes, leading to system response lag and even conflicts between voltage regulation and temperature control. Furthermore, existing solutions lack proactive prediction and compensation mechanisms for temperature change trends, making it difficult to achieve efficient and stable coordinated control under complex operating conditions. Summary of the Invention

[0004] The purpose of this invention is to provide an adaptive temperature control method and transformer system for multi-level power topologies, which realizes dynamic linkage control between voltage output and temperature changes, improves the system's response to load fluctuations and thermal management accuracy, and effectively solves the system instability problem caused by untimely voltage regulation or lag in temperature control in the prior art, thereby solving the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: an adaptive temperature control method for a multi-level power topology, comprising the following steps:

[0006] The real-time temperature of the target device is detected, first temperature data is generated, and based on the first temperature data, the current load status is analyzed and divided into preset multi-level power topology layers;

[0007] Based on the multi-level power topology hierarchy, a voltage adjustment command corresponding to the hierarchy is generated, and a control signal is generated by dynamically matching the voltage adjustment command with the input characteristics of the target device.

[0008] The control signal is used to drive the power topology switching module to adjust the voltage output level, monitor the effect of the adjusted voltage output level on the temperature, and generate second temperature data.

[0009] By comparing the first temperature data with the second temperature data, a temperature compensation value is calculated. Combining the temperature compensation value with the current power topology level, the voltage adjustment command is updated and the next round of topology switching is triggered.

[0010] Preferably, generating the first temperature data includes:

[0011] Collect temperature values ​​at multiple points on the surface of the target device to form an initial temperature sequence;

[0012] The initial temperature sequence is subjected to sliding window mean calculation, and the temperature values ​​within the window are weighted using a weight allocation method to obtain the filtered temperature values.

[0013] The filtered temperature value is compared with the preset reference temperature, the difference between the two is calculated, and the abnormal temperature points are corrected according to the dynamic compensation factor to obtain the corrected temperature value. The corrected temperature value is then output as the first temperature data.

[0014] Preferably, the step of analyzing the current load state and dividing it into preset multi-level power topology layers includes:

[0015] Obtain the current voltage output level and load current sampling value, and calculate the current power consumption value;

[0016] The current power consumption value is compared with the historical average power to obtain the power change rate value;

[0017] Based on the range of the power change rate value, a preset power-level mapping table is matched to determine the corresponding power topology level;

[0018] Based on the power topology level and the current temperature trend, it is determined whether a topology switch needs to be triggered in advance. If the conditions are met, it is marked as a state to be switched.

[0019] Preferably, the generation of voltage adjustment instructions corresponding to the level includes:

[0020] Based on the currently determined power topology level, retrieve the preset voltage offset reference value;

[0021] Based on the maximum allowable voltage fluctuation range corresponding to the power topology level and the current output voltage deviation, the available adjustment range is calculated;

[0022] If the current output voltage deviation exceeds the allowable range, the voltage adjustment range is limited to the boundary value, and a corrected voltage offset is generated.

[0023] The voltage offset reference value is added to the corrected voltage offset to generate the final voltage adjustment command.

[0024] Preferably, the generation control signal includes:

[0025] Receive the voltage adjustment command and the current input voltage, and calculate the voltage difference between the two;

[0026] Based on the voltage difference and the preset slope parameter, determine the time step required for voltage regulation;

[0027] The voltage change process is divided into multiple pulse sequences with equal time intervals, and each segment outputs a corresponding duty cycle value.

[0028] The duty cycle value is converted into a drive pulse signal and output to the power switching element through an isolation circuit to form a continuously adjustable control signal.

[0029] Preferably, the adjusted voltage output level includes:

[0030] It receives drive pulse signals and selects the conduction path through a power switch array to form a primary transformer circuit;

[0031] Based on the conduction path, control the multi-stage winding connection ratio, adjust the transformer ratio, and activate the corresponding synchronous rectification unit on the basis of the transformer.

[0032] The output voltage is monitored by a feedback sampling circuit. If the deviation from the target value exceeds the tolerance, the winding connection ratio is finely adjusted to achieve closed-loop voltage regulation.

[0033] Preferably, generating the second temperature data includes:

[0034] After the voltage output level adjustment is completed, start timed sampling to obtain a new round of multi-point temperature readings of the target device;

[0035] The multi-point temperature readings are processed using an exponentially weighted moving average, and the average temperature value is calculated by combining the attenuation parameter.

[0036] The average temperature value is compared with a preset temperature change threshold. If the difference exceeds the range, it is marked as a significant temperature change event.

[0037] By combining the state of the significant temperature change event with the average temperature value, a second temperature data is generated.

[0038] Preferably, the calculation of the temperature compensation value includes:

[0039] Obtain the first temperature data before adjustment and the second temperature data after adjustment, and calculate the temperature difference between them;

[0040] Based on the temperature difference and the current voltage output level, derive the temperature response corresponding to a unit voltage change;

[0041] Based on the temperature response and the preset reference voltage step size, the expected temperature offset is calculated, and the expected temperature offset is written into the adjustment parameter pool as a temperature compensation value.

[0042] Preferably, the update voltage adjustment command and triggering the next round of topology switching includes:

[0043] Read the latest written temperature compensation value from the adjustment parameter pool and combine it with the reference voltage of the current power topology level;

[0044] Based on the temperature compensation value, the voltage offset is pre-corrected, and the corrected target voltage value is calculated.

[0045] The corrected voltage target value is compared with the output voltage limit range. If it exceeds the boundary, it is truncated to the allowable range, and the final updated voltage adjustment command is generated.

[0046] The power-level mapping table is rematched according to the voltage adjustment command. If the corresponding level is different from the current level, a new round of topology switching request is triggered.

[0047] On the other hand, this invention proposes an adaptive temperature-controlled transformer system with a multi-stage power topology, comprising:

[0048] The load status analysis module is used to detect the real-time temperature of the target device, generate first temperature data, analyze the current load status based on the first temperature data, and classify it into a preset multi-level power topology layer.

[0049] The control signal matching module is used to generate voltage adjustment commands corresponding to the multi-level power topology hierarchy, and generate control signals by dynamically matching the voltage adjustment commands with the input characteristics of the target device.

[0050] The voltage regulation execution module is used to drive the power topology switching module with the control signal, adjust the voltage output level, monitor the effect of the adjusted voltage output level on the temperature, and generate second temperature data.

[0051] The voltage command update module is used to calculate the temperature compensation value by comparing the first temperature data with the second temperature data, and then update the voltage adjustment command and trigger the next round of topology switching by combining the temperature compensation value with the current power topology level.

[0052] Technical effects and advantages of the present invention: The adaptive temperature control method and transformer system for multi-stage power topology proposed in this invention have the following advantages compared with the prior art:

[0053] This invention generates first temperature data by collecting the real-time temperature of the target device and uses this data to classify the power topology level of the current load. Subsequently, it generates a corresponding voltage adjustment command based on this level and generates a control signal matching the input characteristics to drive topology switching. After voltage adjustment, it continues to collect second temperature data, calculates temperature compensation values ​​through comparative analysis, and uses this to update the voltage command, forming a closed-loop control. This achieves dynamic linkage control between voltage output and temperature changes, improving the system's responsiveness to load fluctuations and thermal management accuracy. It effectively solves the system instability problems caused by untimely voltage regulation or lagging temperature control in existing technologies, thereby improving overall operating efficiency and reliability. Attached Figure Description

[0054] Figure 1 This is a flowchart of the adaptive temperature control method for the multi-level power topology of the present invention;

[0055] Figure 2 This is a block diagram of the adaptive temperature control transformer system with multi-stage power topology of the present invention;

[0056] Figure 3 This is a graph showing the temperature correction process of the present invention.

[0057] Figure 4 This is a timing diagram of the voltage regulation process of the present invention. Detailed Implementation

[0058] 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 specific embodiments described herein are merely used to explain the present invention and are not intended to limit the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0059] This invention provides, for example Figure 1 An adaptive temperature control method for a multi-level power topology, as shown, includes the following steps:

[0060] Step 1: Detect the real-time temperature of the target device and generate initial temperature data; this includes the following steps:

[0061] A distributed thermistor array is used to collect temperature values ​​at multiple points on the surface of the target device to form an initial temperature sequence.

[0062] The temperature after filtering is calculated by applying a sliding window mean to the initial temperature sequence, as shown in the formula. in:

[0063] Tfiltered It is a temperature value that has been weighted and averaged to reduce noise and instantaneous fluctuations in the original temperature reading, providing a more stable and smoother temperature representation.

[0064] T i This represents the i-th temperature reading within the sliding window. Here, i refers to each individual temperature measurement point within the window, and each T... i It represents the actual temperature measurement result at a specific point in time.

[0065] W i For corresponding to T i The weighting coefficients reflect the importance of a specific temperature reading in the overall averaging calculation. Different applications may require adjustments to these weights based on the specific circumstances; for example, newer data may be given higher weights to respond more quickly to the latest changes.

[0066] ∑(T i ·W i The sum of the products of all temperature readings within the window and their respective weights is used. This summation ensures that readings considered more important (with higher weights) have a greater impact on the final result.

[0067] ∑W i It is the sum of all weight coefficients. Used as the denominator, it ensures that the result of the weighted average is not affected by the absolute size of the weights, but rather depends on the relative proportions between the individual readings.

[0068] The purpose of the entire formula is to generate a more accurate and stable temperature value (T) by weighting a series of temperature readings. filtered This method helps eliminate the impact of single or occasional abnormal temperature readings on the overall assessment, and allows for the adjustment of the importance of different readings based on specific circumstances (by changing their respective weights W). i This makes the final temperature value more reflective of the actual situation.

[0069] Based on T filtered The difference ΔT between the temperature and the preset reference temperature is used to correct for abnormal temperature points. The correction formula is T. co =T filtered +α*ΔT, where:

[0070] T co At the original filtered temperature (T) filtered Based on the difference from the reference temperature (ΔT) and a dynamic compensation factor (α), the final temperature reading is calculated to eliminate errors caused by outliers or environmental factors.

[0071] α is a dynamic compensation factor, an adjustment coefficient that determines the correction magnitude based on ΔT. The choice of α can be adjusted according to the specific application scenario to better adapt to different environmental conditions or error levels. For example, in situations requiring rapid response to changes, α can be set higher; while in cases where stability is prioritized, a smaller α value may be chosen.

[0072] ΔT is the temperature difference, representing the filtered temperature value (T). filtered The difference between the current temperature reading and the preset reference temperature. This difference is used to assess whether the current temperature reading deviates from the expected range and to make necessary corrections accordingly.

[0073] The overall meaning of this formula is to filter the temperature value (T) filtered The temperature is adjusted proportionally (by α) based on the difference between the temperature and the reference temperature (ΔT) to obtain a more accurate temperature reading (T). co This method is particularly suitable for systems requiring high-precision temperature control, effectively reducing errors caused by external interference or the inherent characteristics of the sensor, and improving system reliability and accuracy. (The last sentence appears to be incomplete and possibly refers to a technical detail about temperature control.) co This is output as the first temperature data and stored in a local buffer for subsequent analysis.

[0074] Step 2: Based on the first temperature data, analyze the current load status and classify it into a preset multi-level power topology hierarchy; including the following steps:

[0075] Get the current voltage output level V output With load current sampling value I sample Calculate the power consumption value P load =V output ·I sample The meaning of this formula is to change the output voltage (V) output ) and load current sampling value (I sample The actual power consumption of the load (P) is calculated by multiplying the two components. load This method provides a direct and effective way to estimate electricity usage at any given point in time.

[0076] P load Compared with historical average power P avg By comparison, the power change rate k = (P) load -P avg ) / P avg k helps identify changes in power consumption patterns, thus providing a basis for system control.

[0077] The specific meaning of this formula is to compare the current load power (P) load ) and historical average power (Pavg The difference between the two values ​​is calculated, and this difference is divided by the historical average power (P). avg This yields a dimensionless proportionality coefficient (k). This proportionality coefficient clearly shows the magnitude and direction (increase or decrease) of the current power consumption relative to the average level.

[0078] If k > 0, it means that the current power consumption is higher than the historical average, and there may be an increase in load.

[0079] If k < 0, it indicates that the current power consumption is lower than the historical average, which may be due to reduced load or improved system efficiency.

[0080] When k≈0, it means that the current power consumption is close to the historical average level and the system is operating stably.

[0081] Based on the range of k, a preset power-level mapping table is matched to determine the corresponding power topology level L; based on L and the current temperature trend, it is determined whether topology switching needs to be triggered in advance, and if the conditions are met, it is marked as a state to be switched.

[0082] Step 3: Generate voltage adjustment commands corresponding to the multi-level power topology hierarchy; including the following steps:

[0083] Based on the currently determined power topology level L, retrieve the preset voltage offset reference value.

[0084] Based on the maximum allowable voltage fluctuation range corresponding to L. Combined with the current output voltage deviation ΔV current =V actual -V target Calculate the available adjustment range; V actual Indicates the current actual output voltage value; V target This indicates the preset target voltage value;

[0085] If ΔV current Exceeding The voltage adjustment range will then be limited to... Generate the corrected voltage offset V offset =clamp(ΔV) current ,-ΔV max ,ΔV max );in:

[0086] `clamp(x,min,max)` is a clamping function that limits the input value `x` to a specified range [min,max]. If `x` is less than `min`, it returns `min`; if `x` is greater than `max`, it returns `max`; if `x` is between `min` and `max`, it returns `x` itself. This process ensures that the output value will not exceed the preset upper and lower limits.

[0087] ΔV current The current output voltage deviation refers to the difference between the actual output voltage and the target voltage. This value reflects the amount of voltage adjustment needed to bring the actual voltage closer to the target voltage.

[0088] -ΔV max ,ΔV max These are the maximum permissible negative and positive voltage fluctuation ranges determined based on the current power topology level L. These two values ​​define the maximum magnitude of voltage adjustment that can be achieved without compromising system performance or safety.

[0089] The meaning of this formula is to use the clamping function to reduce the current output voltage deviation (ΔV) current ) limited to the maximum allowable voltage fluctuation range ( arrive Within the range defined by (), this is done to prevent excessive voltage adjustment from causing system instability or other potential problems. Specifically:

[0090] If ΔV current Less than Then V offset Set as

[0091] If ΔV current Greater than Then V offset Set as

[0092] If ΔV current exist arrive Between, then V offset Directly equal to ΔV current .

[0093] This method ensures that any voltage adjustments remain within the system's safe operating range, helping to maintain system stability and reliability and avoiding problems caused by over-adjustment. With V offset Add them together to generate the final voltage adjustment command.

[0094] Step 4: Generate a control signal by dynamically matching the voltage adjustment command with the input characteristics of the target device; this includes the following steps:

[0095] Receive voltage adjustment command V command With the current input voltage V input Calculate the difference between the two, ΔV = V command -V input Based on ΔV and a preset slope factor S, the time step N required for voltage regulation is determined as N = abs(ΔV) / S. This formula calculates the total number of steps N required to complete voltage regulation by dividing the absolute value of the voltage difference by the maximum allowable rate of change. Where:

[0096] ΔV is the voltage difference, i.e., the difference between the target voltage and the current actual voltage; abs indicates taking its absolute value, ensuring that N is a positive number regardless of whether the voltage increases or decreases. S represents the maximum allowable voltage change per unit time. It reflects the rate of change that the system can withstand during voltage regulation, and is usually determined by hardware responsiveness or system stability requirements.

[0097] Based on N, the voltage change process is divided into multiple pulse sequences with equal time intervals, and the duty cycle of each segment is D. i = (i*ΔV) / N, where i is an integer from 1 to N, representing the stage into which the entire voltage regulation process is divided.

[0098] The formula calculates the voltage increment percentage to be achieved in each stage by multiplying the current segment number i by the total voltage difference ΔV and dividing by the total number of steps N, and then converts this percentage into the corresponding duty cycle value D. i This allows for segmented control of the voltage change process, enabling the voltage to gradually and smoothly transition from the current value to the target value.

[0099] As i increases (i.e., when processing later paragraphs), D i This also increases accordingly, indicating that the conduction time of the medium power switch in each segment gradually increases;

[0100] After the entire process is completed, the voltage reaches the target value exactly, avoiding the impact caused by sudden large jumps. This method is often used in digital control power supplies or PWM (pulse width modulation) systems to achieve precise and continuous regulation of the output voltage.

[0101] D i It is converted into a drive pulse signal and output to the power switching element through an isolation circuit to form a continuously adjustable control signal.

[0102] Step 5: Use the control signal to drive the power topology switching module and adjust the voltage output level; this includes the following steps:

[0103] It receives drive pulse signals and selects the conduction path through a power switch array to form a primary transformer circuit;

[0104] Based on the conduction path control, the multi-stage winding connection ratio n ratio =N primary / N secondary Adjust the transformer turns ratio;

[0105] N primary Primary winding turns refer to the total number of turns in the primary winding of a transformer. This is the part connected to the power supply side, and its number of turns determines how the input voltage is transmitted to the core, further affecting the output on the secondary side.

[0106] N secondary The number of turns in the secondary winding refers to the total number of turns in the secondary winding of the transformer. This part of the winding is responsible for converting the energy transmitted through the magnetic core into the required output voltage.

[0107] Based on the transformer, the corresponding synchronous rectification unit is activated to improve the energy transmission efficiency η = P. out / P in η reflects the efficiency of the system in energy conversion or transmission.

[0108] P out Output power refers to the actual useful power that a system or device outputs to the load.

[0109] P in Input power refers to the total power supplied to the system from the power source. The output voltage V is monitored by a feedback sampling circuit. out If the deviation from the target value exceeds the tolerance limit, then n is fine-tuned. ratio Achieve closed-loop voltage regulation.

[0110] Step Six: Monitor the impact of the adjusted voltage output level on temperature and generate secondary temperature data; this includes the following steps:

[0111] After the voltage output level adjustment is completed, the timed sampling mechanism is activated to obtain a new round of multi-point temperature readings of the target device;

[0112] The multi-point temperature readings are processed using an exponentially weighted moving average, and the calculation formula is T. ewma =β·T current +(1-β)·T prev ,in:

[0113] T ewma Exponentially weighted moving average temperature value. This is a processed temperature value that reflects not only the currently measured temperature (T) current It also incorporates historical temperature data (T). prevThis is used to smooth out noise in temperature readings and identify trends in temperature changes.

[0114] β: Attenuation factor or weighting coefficient. This value determines the relative importance of new and old temperature data when calculating the exponentially weighted moving average. β typically ranges from 0 to 1. When β is closer to 1, more weight is given to the most recent temperature reading (T). current The influence of historical data is relatively small; conversely, if β is small, it indicates that past temperature data is given more importance, resulting in a smoother change in temperature trends.

[0115] T current : The temperature reading collected at the current moment. This is the latest temperature measurement result, directly reflecting the actual temperature status of the target device at this point in time.

[0116] T prev The previously calculated exponentially weighted moving average temperature value (i.e., T for the previous period) ewma By combining it with the current temperature reading, this ensures that the temperature estimate can quickly respond to changes in actual temperature without large jumps due to abnormal fluctuations in a single reading. This formula is mainly used in the process of generating secondary temperature data, by performing an exponentially weighted moving average on multiple temperature readings to obtain a more stable and reliable temperature representation.

[0117] T ewma Compared with the preset temperature change threshold ΔT th Compare the values; if the difference exceeds the range, mark it as a significant temperature change event.

[0118] Combining the state of significant temperature change events with T ewma The second temperature data is generated and stored in the historical record area for the next round of compensation calculation.

[0119] Step 7: Calculate the temperature compensation value by comparing the first temperature data with the second temperature data; this includes the following steps:

[0120] Obtain the first temperature data T before adjustment initial Compared with the adjusted second temperature data T final Calculate the temperature difference ΔT adj =T final -T initial Based on ΔT adj and the current voltage output level V output The temperature response δ = ΔT corresponding to a unit voltage change is derived. adj / (V final -V initial ); where V final -V initialThis indicates that the output voltage during this adjustment is from the initial value (V). initial ) to the final value (V) final The difference between the two values. This difference can be positive (boost) or negative (blow down).

[0121] If δ is large, it indicates that the system is very sensitive to voltage changes; even small voltage adjustments can cause significant temperature changes, requiring careful adjustment. If δ is small, it indicates that the system is relatively stable, and voltage fluctuations have a limited impact on temperature. This is determined by δ and the preset reference voltage step size ΔV. ref Calculate the expected temperature offset ΔT est =δ*ΔV ref ; will ΔT est The temperature compensation value is written into the adjustment parameter pool and used as the basis for the pre-correction when generating the next round of voltage commands.

[0122] Step 8: Combining the temperature compensation value with the current power topology level, update the voltage adjustment command and trigger the next round of topology switching; including the following steps:

[0123] Read the latest written temperature compensation value ΔT from the adjustment parameter pool. est And combined with the reference voltage of the current power topology level L

[0124] Based on ΔT est Perform a pre-correction on the voltage offset and calculate the corrected target voltage value. Where K is the mapping proportionality constant; it represents the proportionality coefficient of the effect of temperature change on voltage regulation. This coefficient reflects how much voltage regulation is needed to compensate for or adapt to temperature changes in a specific application environment. It can be adjusted according to the specific application scenario and experimental data.

[0125] V new The voltage is compared with the output voltage limit range. If it exceeds the boundary, it is truncated to the allowable range, and the final updated voltage adjustment command is generated.

[0126] The power-level mapping table is rematched based on the updated voltage adjustment instructions. If the corresponding level is different from the current level, a new round of topology switching requests is triggered.

[0127] On the other hand, this invention proposes an adaptive temperature-controlled transformer system with a multi-stage power topology, comprising:

[0128] The load status analysis module is used to detect the real-time temperature of the target device, generate first temperature data, analyze the current load status based on the first temperature data, and classify it into a preset multi-level power topology layer.

[0129] The control signal matching module is used to generate voltage adjustment commands corresponding to the multi-level power topology hierarchy, and generate control signals by dynamically matching the voltage adjustment commands with the input characteristics of the target device.

[0130] The voltage regulation execution module is used to drive the power topology switching module with the control signal, adjust the voltage output level, monitor the effect of the adjusted voltage output level on the temperature, and generate second temperature data.

[0131] The voltage command update module is used to calculate the temperature compensation value by comparing the first temperature data with the second temperature data, and then update the voltage adjustment command and trigger the next round of topology switching by combining the temperature compensation value with the current power topology level.

[0132] In addition, the modules mentioned above are also used to implement other steps of the adaptive temperature control method for a multi-level power topology, as shown in the following example:

[0133] Assuming the target device is a high-power server, it will generate significant heat during operation. The system needs to dynamically adjust the voltage output based on the real-time temperature to avoid overheating or energy waste.

[0134] Step 1: Temperature Acquisition and Correction

[0135] A distributed thermistor array collects the temperature at three points: T1 = 42℃, T2 = 45℃, T3 = 43℃; the sliding window weight W = [0.2, 0.5, 0.3].

[0136] Calculate the weighted average temperature: T filtered =(T1*W1+T2*W2+T3*W3) / (W1+W2+W3)

[0137] = (42*0.2+45*0.5+43*0.3) / (0.2+0.5+0.3)

[0138] = (8.4 + 22.5 + 12.9) / 1.0 = 43.8℃;

[0139] Correcting abnormal temperature points, reference temperature T base =40℃, ΔT=43.8-40=3.8℃, dynamic compensation factor α=0.3;

[0140] T co =T filtered +α*ΔT=43.8+0.3*3.8=43.8+1.14=44.94℃;

[0141] First temperature data T initial =44.94℃.

[0142] Step 2: Load Analysis and Hierarchical Division

[0143] Calculate the current power P load Output voltage V output =12V, current I sample =8A;

[0144] Formula: P load =V output ·I sample =12*8=96W;

[0145] Calculate the rate of change of power k: historical average power P avg =80W, formula:

[0146] k=(P load -P avg ) / P avg = (96-80) / 80 = 0.2;

[0147] k = 0.2 (higher than the baseline value of 0.1, indicating an increase in load).

[0148] Matching power-hierarchy mapping table:

[0149] Assume the mapping rule: k∈[0.1,0.3]→level L=2 (medium load);

[0150] The current temperature trend (44.94℃) has not exceeded the threshold and will not trigger an early switch.

[0151] Step 3: Generate voltage adjustment command

[0152] Retrieve the reference voltage at level L=2 Preset table

[0153] Calculate voltage deviation and limiting:

[0154] Actual voltage V actual =11.2V, target voltage V target =11.5V;

[0155] Formula: ΔV current =V actual -V target =11.2-11.5=-0.3V;

[0156] Maximum fluctuation range of level L=2

[0157] Formula: V offset =clamp(ΔV) current -0.5, 0.5) = -0.3V (not exceeding the limit).

[0158] Generate final voltage command: Formula:

[0159] Step 4: Generate control signals

[0160] Calculate the voltage regulation step size N:

[0161] Current input voltage V input =11.0V, ΔV=V command -V input =0.2V;

[0162] Slope factor S = 0.1V / step;

[0163] Formula: N = abs(ΔV) / S = 0.2 / 0.1 = 2 steps.

[0164] Generate duty cycle sequence D i The formulas are: D1 = (1 * 0.2) / 2 = 0.1; D2 = (2 * 0.2) / 2 = 0.2;

[0165] The duty cycle sequence [0.1, 0.2] is converted into a PWM signal to drive the power switch.

[0166] Step 5: Adjust the voltage output level

[0167] Switching power topology:

[0168] Number of turns in primary winding N primary =100, number of secondary turns N secondary =80;

[0169] Formula: n ratio =N primary / N secondary =100 / 80=1.25;

[0170] After adjusting the transformer turns ratio, the output voltage stabilized at 11.2V.

[0171] Synchronous rectification efficiency calculation: Output power P out =11.2V*8A=89.6W; Input power P in =12V * 8A = 96W; Formula: η = P out / P in =89.6 / 96≈0.933 (93.3%).

[0172] Step Six: Generate Second Temperature Data

[0173] Temperature after adjustment:

[0174] New round of temperature readings: T1 = 43℃, T2 = 44℃, T3 = 43℃;

[0175] Use an exponentially weighted moving average (β = 0.7);

[0176] formula:

[0177] T ewma =β·T current +(1-β)·T prev =0.7*43 + 0.3*44.94 = 30.1 + 13.48 = 43.58℃; Second temperature data T final =43.58℃.

[0178] like Figure 3 The temperature correction process curve shown in the figure demonstrates that the system successfully reduced the temperature from the initial 45℃ to 43.58℃, and the dynamic compensation stage briefly raised the temperature to 44.94℃, verifying the effectiveness of the compensation factor α = 0.3.

[0179] Step 7: Calculate the temperature compensation value

[0180] Calculate the temperature difference ΔT_adj using the formula: ΔT adj =T final -T initial =43.58-44.94=-1.36℃;

[0181] Derivation of the unit voltage temperature response δ: voltage change V final -V initial =11.2V - 11.5V = -0.3V;

[0182] Formula: δ=ΔT adj / (V final -V initial )=-1.36 / (-0.3)≈4.53℃ / V;

[0183] Calculate the expected temperature offset ΔT est Reference voltage step size ΔV ref =0.2V, formula: ΔT est =δ*ΔV ref =4.53 * 0.2 ≈ 0.91℃;

[0184] Step 8: Update instructions and trigger the switch.

[0185] Corrected voltage target value V new : Current level L=2 Mapping scaling constant K = 0.1; Formula:

[0186] Limiting: Output voltage limiting range [11.0V, 12.0V], V new =11.59V compliant;

[0187] Rematch tier: Update voltage command to 11.59V; match power-tier mapping table according to the new command, tier remains L=2, no switching required.

[0188] like Figure 4 The timing diagram of the voltage regulation process shown shows that the system achieves precise voltage regulation: there is a 0.3V deviation between the initial voltage (11.2V) and the target voltage (11.5V). After compensation, the final output stabilizes at 11.59V. The entire process complies with the ±0.5V fluctuation limit of L=2 level.

[0189] In this example, by dynamically adjusting the voltage, the temperature dropped from 44.94℃ to 43.58℃, effectively suppressing overheating. The synchronous rectification efficiency reached 93.3%, enhancing system stability. Temperature compensation value ΔT est =0.91℃ was used for the next round of voltage command correction, resulting in continuous optimization.

[0190] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. An adaptive temperature control method for a multi-level power topology, characterized in that, Includes the following steps: The real-time temperature of the target device is detected, first temperature data is generated, and based on the first temperature data, the current load status is analyzed and divided into preset multi-level power topology layers; Based on the multi-level power topology hierarchy, a voltage adjustment command corresponding to the hierarchy is generated, and a control signal is generated by dynamically matching the voltage adjustment command with the input characteristics of the target device. The control signal is used to drive the power topology switching module to adjust the voltage output level, monitor the effect of the adjusted voltage output level on the temperature, and generate second temperature data. By comparing the first temperature data with the second temperature data, a temperature compensation value is calculated. Combining the temperature compensation value with the current power topology level, the voltage adjustment command is updated and the next round of topology switching is triggered. The generation of voltage adjustment instructions corresponding to the level includes: Based on the currently determined power topology level, retrieve the preset voltage offset reference value; Based on the maximum allowable voltage fluctuation range corresponding to the power topology level and the current output voltage deviation, the available adjustment range is calculated; If the current output voltage deviation exceeds the allowable range, the voltage adjustment range is limited to the boundary value, and a corrected voltage offset is generated. The voltage offset reference value is added to the corrected voltage offset to generate the final voltage adjustment command; The generation control signal includes: Receive the voltage adjustment command and the current input voltage, and calculate the voltage difference between the two; Based on the voltage difference and the preset slope parameter, determine the time step required for voltage regulation; The voltage change process is divided into multiple pulse sequences with equal time intervals, and each segment outputs a corresponding duty cycle value. The duty cycle value is converted into a drive pulse signal and output to the power switching element through an isolation circuit to form a continuously adjustable control signal. The calculation of the temperature compensation value includes: Obtain the first temperature data before adjustment and the second temperature data after adjustment, and calculate the temperature difference between them; Based on the temperature difference and the current voltage output level, derive the temperature response corresponding to a unit voltage change; Based on the temperature response and the preset reference voltage step size, the expected temperature offset is calculated, and the expected temperature offset is written into the adjustment parameter pool as a temperature compensation value. The update voltage adjustment command and triggering of the next round of topology switching include: Read the latest written temperature compensation value from the adjustment parameter pool and combine it with the reference voltage of the current power topology level; Based on the temperature compensation value, the voltage offset is pre-corrected, and the corrected target voltage value is calculated. The corrected voltage target value is compared with the output voltage limit range. If it exceeds the boundary, it is truncated to the allowable range, and the final updated voltage adjustment command is generated. The power-level mapping table is rematched according to the voltage adjustment command. If the corresponding level is different from the current level, a new round of topology switching request is triggered.

2. The adaptive temperature control method for a multi-stage power topology according to claim 1, characterized in that: The generation of the first temperature data includes: Collect temperature values ​​at multiple points on the surface of the target device to form an initial temperature sequence; The initial temperature sequence is subjected to sliding window mean calculation, and the temperature values ​​within the window are weighted using a weight allocation method to obtain the filtered temperature values. The filtered temperature value is compared with the preset reference temperature, the difference between the two is calculated, and the abnormal temperature points are corrected according to the dynamic compensation factor to obtain the corrected temperature value. The corrected temperature value is then output as the first temperature data.

3. The adaptive temperature control method for a multi-stage power topology according to claim 2, characterized in that: The analysis of the current load status and its classification into preset multi-level power topology layers includes: Obtain the current voltage output level and load current sampling value, and calculate the current power consumption value; The current power consumption value is compared with the historical average power to obtain the power change rate value; Based on the range of the power change rate value, a preset power-level mapping table is matched to determine the corresponding power topology level; Based on the power topology level and the current temperature trend, it is determined whether a topology switch needs to be triggered in advance. If the conditions are met, it is marked as a state to be switched.

4. The adaptive temperature control method for a multi-stage power topology according to claim 1, characterized in that: The adjusted voltage output level includes: It receives drive pulse signals and selects the conduction path through a power switch array to form a primary transformer circuit; Based on the conduction path, control the multi-stage winding connection ratio, adjust the transformer ratio, and activate the corresponding synchronous rectification unit on the basis of the transformer. The output voltage is monitored by a feedback sampling circuit. If the deviation from the target value exceeds the tolerance, the winding connection ratio is finely adjusted to achieve closed-loop voltage regulation.

5. The adaptive temperature control method for a multi-stage power topology according to claim 3, characterized in that: The generation of the second temperature data includes: After the voltage output level adjustment is completed, start timed sampling to obtain a new round of multi-point temperature readings of the target device; The multi-point temperature readings are processed using an exponentially weighted moving average, and the average temperature value is calculated by combining the attenuation parameter. The average temperature value is compared with a preset temperature change threshold. If the difference exceeds the range, it is marked as a significant temperature change event. By combining the state of the significant temperature change event with the average temperature value, a second temperature data is generated.

6. An adaptive temperature-controlled transformer system for implementing a multi-stage power topology as described in any one of claims 1-5, characterized in that, include: The load status analysis module is used to detect the real-time temperature of the target device, generate first temperature data, analyze the current load status based on the first temperature data, and classify it into a preset multi-level power topology layer. The control signal matching module is used to generate voltage adjustment commands corresponding to the multi-level power topology hierarchy, and generate control signals by dynamically matching the voltage adjustment commands with the input characteristics of the target device. The voltage regulation execution module is used to drive the power topology switching module with the control signal, adjust the voltage output level, monitor the effect of the adjusted voltage output level on the temperature, and generate second temperature data. The voltage command update module is used to calculate the temperature compensation value by comparing the first temperature data with the second temperature data, and then update the voltage adjustment command and trigger the next round of topology switching by combining the temperature compensation value with the current power topology level.

Images