High-frequency transformer with automatic heat dissipation function

Abstract

This invention belongs to the field of transformer technology, and specifically relates to a high-frequency transformer with automatic heat dissipation function, including a transformer core, a wound coil, a fixing frame, a piezoelectric fan, a semiconductor cooling chip, a heat transfer plate, and a control module. The wound coil is wound on the transformer core, the fixing frame covers the transformer core and the wound coil, the semiconductor cooling chip is fixedly installed on the outer wall of the heat transfer plate, and the piezoelectric fan is installed near the hot end of the semiconductor cooling chip. The control module is electrically connected to the semiconductor cooling chip and the piezoelectric fan. The control module is configured to: acquire the current and voltage signals of the wound coil in real time, calculate the real-time heating power based on the current and voltage signals and output a basic cooling control signal accordingly, and calculate the time change rate of dynamic resistance based on the same current and voltage signals. This eliminates the heat conduction delay of traditional temperature feedback, and the response speed reaches the microsecond level, making it particularly suitable for high-temperature environments and precision control scenarios.

Description

Technical Field

[0001] This invention belongs to the field of transformer technology, specifically a high-frequency transformer with automatic heat dissipation function. Background Technology

[0002] High-frequency transformers are widely used in switching power supplies, inverters, communication equipment, and precision instruments. With increasing power density and diversified application environments, the heat dissipation problem of transformers is becoming increasingly prominent. On the one hand, in high-temperature environments, passive cooling alone is insufficient to effectively dissipate the heat generated by the winding coils and transformer core. On the other hand, in precision control environments, even small fluctuations in the temperature of the winding coils can cause changes in electrical parameters; therefore, it is necessary to control the temperature within a narrow range and to respond quickly to sudden load changes.

[0003] To address these needs, existing technologies employ direct cooling with semiconductor thermoelectric coolers and feedback control via temperature sensors. However, temperature sensors are typically mounted on the surface of the wound coil or a fixed frame. Heat transfer from the internal heating point to the sensor requires traversing multiple layers of media, resulting in significant measurement delays. When the load suddenly increases, the internal temperature of the wound coil may have already risen before the sensor detects it, leading to a lag in controller response and temperature overshoot. While embedding the sensor inside the wound coil shortens the heat transfer path, it damages the insulation layer and introduces electromagnetic interference. Furthermore, the accuracy and stability of ordinary sensors decrease in high-temperature environments.

[0004] Another known approach is to estimate the temperature by measuring the DC resistance of the wound coil, which is used to implement overheat alarm or trip protection. This approach is based on the physical principle that resistance changes with temperature, but it is only used for protection purposes. Its calculation is based on a steady-state model, and the resistance signal processing is a threshold comparison on a long time scale. The output is a switching signal. However, the rate of resistance change itself implies the temperature change trend information. Existing protection schemes do not utilize this information, nor do they use it for the continuous dynamic adjustment of active heat dissipation devices.

[0005] Therefore, the present invention provides a high-frequency transformer with automatic heat dissipation function. Summary of the Invention

[0006] In order to overcome the shortcomings of the prior art, at least one technical problem raised in the background art is solved.

[0007] The technical solution adopted by this invention to solve its technical problem is: a high-frequency transformer with automatic heat dissipation function, comprising... Transformer core, A wound coil, the wound coil being wound on the transformer core; A fixed frame, which covers the outside of the transformer core and the winding coil; A heat transfer plate, which is mounted on the fixed frame; A semiconductor refrigeration chip, wherein the semiconductor refrigeration chip is fixedly mounted on the outer wall of a heat transfer plate; A piezoelectric fan, wherein the piezoelectric fan is mounted near the hot end of the semiconductor cooling chip and directly facing the hot end; A control module, which is electrically connected to both the thermoelectric cooler and the piezoelectric fan, is configured to: The system acquires the current and voltage signals of the wound coil in real time, calculates the real-time heat generation power based on the current and voltage signals, and outputs the basic cooling control signal accordingly. It also calculates the time change rate of dynamic resistance based on the same current and voltage signals, generates a feedforward correction signal, and superimposes it onto the basic cooling control signal. Furthermore, it calculates the efficiency deviation of the heat dissipation system based on the real-time heat generation power and the time change rate of dynamic resistance, and outputs a warning signal when the efficiency deviation exceeds a threshold.

[0008] Preferably, when the control module calculates the real-time heating power, it first calculates the active power and the effective value of the current from the current signal and the voltage signal, calculates the dynamic resistance of the winding coil based on the active power and the effective value of the current, then calculates the copper loss using the dynamic resistance, and calculates the iron loss by combining the loss model of the transformer core. The sum of the copper loss and the iron loss is the real-time heating power.

[0009] Preferably, the control module also calculates the real-time temperature or temperature change rate of the wound coil based on the dynamic resistance and resistance-temperature model. The resistance temperature model is: The real-time resistance of the wound coil is equal to the resistance value at the reference temperature multiplied by one, plus the resistance temperature coefficient multiplied by the temperature difference, where the resistance value at the reference temperature and the resistance temperature coefficient are pre-stored constants. The loss model of the transformer core is realized through a pre-calibrated voltage, frequency and iron loss mapping table.

[0010] Preferably, when the control module calculates the time-varying rate of change of the dynamic resistance, it first considers the active power. and the effective value of current Calculate dynamic resistance Calculation formula: The rate of change of resistance is then obtained by differentiating the dynamic resistance over time. Finally, the temperature change rate of the wound coil is calculated based on the temperature coefficient of resistance and the reference resistance value.

[0011] Preferably, before calculating the dynamic resistance, the control module first deducts the fixed inductive reactance component of the wound coil to eliminate the influence of inductance on the resistance calculation.

[0012] Preferably, the feedforward correction signal is generated by multiplying the temperature change rate of the winding coil by a preset first gain coefficient and superimposing it onto the driving voltage of the basic cooling control signal used to control the semiconductor cooling chip. At the same time, the temperature change rate of the winding coil is multiplied by a preset second gain coefficient and superimposed on the driving voltage amplitude of the piezoelectric fan in the basic cooling control signal.

[0013] Preferably, the efficiency deviation of the heat dissipation system is calculated as follows: the expected temperature rise rate is calculated based on the real-time heat dissipation power, the actual heat dissipation power of the current semiconductor cooling chip and piezoelectric fan, and the heat capacity of the transformer body. The actual temperature rise rate obtained from the dynamic resistance change rate over time is then subtracted from the expected temperature rise rate to obtain the efficiency deviation; when the efficiency deviation is greater than a preset threshold, the heat dissipation system is judged to be in abnormal efficiency.

[0014] Preferably, the driving frequency of the piezoelectric fan is fixed to the first-order resonant frequency of its blades, and the control module continuously changes the amplitude of the blades by adjusting the amplitude of the driving voltage, thereby adjusting the air volume.

[0015] Preferably, after determining that the efficiency deviation exceeds the threshold, the control module also performs at least one of the following operations: automatically reducing the maximum allowable input power of the transformer; Increase the driving voltage amplitude of the piezoelectric fan to compensate for the performance degradation of the thermoelectric cooler.

[0016] The beneficial effects of this invention are as follows: 1. This invention calculates the heating power by acquiring the current and voltage signals of the wound coil in real time, eliminating the need for a temperature sensor, thus eliminating the thermal conduction delay of traditional temperature feedback. The response speed reaches the microsecond level, making it particularly suitable for high-temperature environments with high current and precision control scenarios.

[0017] 2. This invention utilizes the time-varying rate of dynamic resistance to convert it into the temperature change rate of the wound coil, generating a feedforward correction signal that is superimposed on the basic cooling signal. This can enhance cooling and airflow in advance when the load changes abruptly, significantly suppress temperature overshoot, and control the temperature fluctuation of the wound coil within an extremely narrow range.

[0018] 3. This invention diagnoses the efficiency deviation of the heat dissipation system in real time by comparing the actual temperature rise rate with the expected temperature rise rate. It can provide early warning of heat dissipation performance degradation before the temperature exceeds the limit and automatically perform remedial operations such as reducing power or increasing fan speed to automatically carry out cooling measures, thereby achieving automatic cooling function, effectively preventing thermal runaway and extending equipment life. Attached Figure Description

[0019] The invention will now be further described with reference to the accompanying drawings.

[0020] Figure 1This is a schematic diagram of the overall structure of the present invention; Figure 2 This is a schematic diagram of the fixed frame structure in this invention; Figure 3 This is a schematic diagram of the heat transfer plate structure in this invention; Figure 4 This is a flowchart of the entire invention.

[0021] In the diagram: 1. Transformer core; 2. Winding coil; 3. Fixing frame; 4. Piezoelectric fan; 5. Semiconductor cooling chip; 6. Heat transfer plate; 7. Base. Detailed Implementation

[0022] To make the technical means, creative features, objectives and effects of this invention easier to understand, the invention will be further described below in conjunction with specific embodiments.

[0023] Example 1: As Figures 1 to 4 As shown in the figure, a high-frequency transformer with automatic heat dissipation function according to an embodiment of the present invention includes a transformer core 1, a wound coil 2, a fixed frame 3, a piezoelectric fan 4, a semiconductor cooling chip 5, a heat transfer plate 6, and a base 7.

[0024] The winding coil 2 is wound on the transformer core 1. The fixing frame 3 covers the outside of the transformer core 1 and the winding coil 2, providing structural support and protection for the transformer. The semiconductor cooling chip 5 is fixedly installed on the outer wall of the heat transfer plate 6 to absorb heat from the inside of the transformer. The outer wall of the heat transfer plate 6 is in contact with the fixing frame 3, and the contact area is evenly coated with silicone grease. The heat transfer plate 6 is made of hollow material and filled with coolant. The cold end of the semiconductor cooling chip 5 is located in the coolant, and its hot end is located outside the heat transfer plate 6.

[0025] The piezoelectric fan 4 is installed near and directly opposite the hot end of the thermoelectric cooler 5, thereby enabling the piezoelectric fan 4 to effectively dissipate the heat emitted by the thermoelectric cooler 5 to the surrounding environment.

[0026] It also includes a control module, which is electrically connected to the semiconductor cooling chip 5 and the piezoelectric fan 4 respectively. The control module is configured to intelligently control the heat dissipation system. The control module can be a microcontroller unit (MCU) that integrates an analog-to-digital converter (ADC), a digital-to-analog converter (DAC), and various communication interfaces to realize signal acquisition, data processing, and control output.

[0027] Specifically, the control module is configured to acquire the current and voltage signals of the winding coil 2 in real time. These signals are a direct reflection of the transformer's operating status. The current signal can be obtained by connecting a shunt resistor in series in the main circuit or by using a current transformer. The voltage signal can be obtained directly from both ends of the winding coil 2 through a voltage divider circuit. The acquired analog signals are then sent to the ADC of the control module for digital processing.

[0028] Based on the acquired current and voltage signals, the control module is configured to calculate the real-time heating power and output a basic cooling control signal accordingly. The real-time heating power can be calculated in various ways. For example, the active power can be estimated by averaging the product of instantaneous current and instantaneous voltage, and this averaged value can be used as an approximation of the real-time heating power.

[0029] Therefore, the basic cooling control signal can be a voltage or current signal proportional to the heat generation power, used to drive the semiconductor cooling chip 5 and the piezoelectric fan 4 to provide basic heat dissipation capabilities.

[0030] In addition, the control module is configured to calculate the time-varying rate of dynamic resistance based on the same current and voltage signals, and generate a feedforward correction signal that is superimposed on the basic cooling control signal. The time-varying rate of dynamic resistance reflects the instantaneous temperature change trend of the wound coil 2.

[0031] The time-varying rate of change of the dynamic resistance reflects the instantaneous temperature change trend of the wound coil 2. Subsequently, by continuously sampling the dynamic resistance value and performing differential operations, its time-varying rate can be obtained. The feedforward correction signal can be a voltage or current increment proportional to the time-varying rate. When a rapid increase in resistance is detected, this increment signal can enhance the cooling capacity of the heat dissipation system in advance to cope with the upcoming temperature rise.

[0032] In a preferred embodiment, the control module is also configured to calculate the efficiency deviation of the heat dissipation system based on the time-varying rate of change of real-time heat generation power and dynamic resistance, and to output a warning signal when the efficiency deviation exceeds a threshold.

[0033] The efficiency deviation of the heat dissipation system can reflect whether the actual working state of the heat dissipation device matches the expectation. The efficiency deviation can be calculated by comparing the theoretical temperature rise rate estimated by the real-time heat generation power with the actual temperature rise rate reflected by the dynamic resistance time change rate. When the deviation value exceeds the preset threshold, it indicates that the heat dissipation system may have performance degradation or failure. The control module then outputs a warning signal, such as by lighting an indicator light, sending a communication message or triggering a buzzer.

[0034] By directly utilizing the current and voltage signals of the wound coil 2 for real-time calculation of heating power and analysis of dynamic resistance time change rate, rapid and delay-free sensing of the thermal state of the high-frequency transformer is achieved. As a result, the heat dissipation system can respond quickly to sudden load changes, effectively suppress temperature overshoot, and avoid the measurement delay, insulation damage, and electromagnetic interference problems caused by traditional temperature sensors. At the same time, through real-time monitoring of the efficiency deviation of the heat dissipation system, this application can promptly detect and warn of system anomalies, ensuring the stable operation and long life of the transformer under complex operating conditions.

[0035] In addition, the control module calculates the real-time heat generation power to output the basic cooling control signal. However, in this process, since the basic calculation is based solely on the current and voltage signals without considering the temperature change of the resistance of the winding coil 2 and the loss characteristics of the transformer core 1, the heat generation power may be inaccurately estimated, which in turn affects the accuracy and response efficiency of the heat dissipation control.

[0036] To address this, the control module first calculates the active power and RMS current value from the current and voltage signals when calculating the real-time heating power. Then, it calculates the dynamic resistance of the wound coil 2 based on the active power and RMS current value. The calculation formula is as follows: The copper loss can be directly calculated using this dynamic resistance, i.e., the copper loss. Meanwhile, the iron loss is calculated using the loss model of transformer core 1. Finally, the sum of copper loss and iron loss is used as the real-time heating power. It should be noted that the resistance-temperature model of winding coil 2 is not used to calculate copper loss, but rather to deduce the real-time temperature or temperature change rate of winding coil 2 from the dynamic resistance. The specific method will be detailed later.

[0037] Specifically, the step of the control module to calculate the real-time heat generation power aims to accurately quantify the heat generated by the high-frequency transformer during operation, providing accurate input for the control of the heat dissipation system. The accuracy of the real-time heat generation power directly affects the timeliness and effectiveness of heat dissipation control, and is the key to avoiding transformer overheating and ensuring its stable operation. This calculation task can be achieved in a variety of ways.

[0038] The control module can use an embedded processor or a digital signal processor (DSP) to perform complex calculation tasks. It can be programmed to sample, process, and calculate the power of current and voltage signals in real time. Alternatively, it can use a dedicated power measurement chip or integrated circuit to convert analog signals into digital signals, perform power calculations, and then transmit the calculation results to the main controller.

[0039] When calculating active power, this step is the basis for calculating the real-time heating power of the transformer. Active power represents the portion of power actually consumed in the circuit and converted into heat energy. Calculating active power by directly measuring electrical parameters can avoid the delays and errors caused by indirect measurements.

[0040] The control module can acquire the voltage signal at both ends of the winding coil 2 and the current signal flowing through the winding coil 2 in real time. It calculates the active power using standard power measurement algorithms such as the instantaneous power integration method or the two-watt meter method. The active power is obtained by integrating the product of the instantaneous voltage and the instantaneous current over one cycle.

[0041] Another approach is to use a dedicated power analyzer chip or module, which can automatically sample voltage and current, perform phase compensation, and calculate active power, and provide the results to the control module via a digital interface (such as SPI or I2C).

[0042] To calculate copper loss more accurately, this application incorporates a resistance-temperature model of the wound coil 2. Copper loss is the main component of the resistive heating of the wound coil 2, and its magnitude is proportional to the square of the resistance and current of the wound coil 2. Since the resistance of the wound coil 2 changes with temperature, calculating copper loss using a resistance-temperature model can more accurately reflect the actual heating of the wound coil 2 and improve the accuracy of the heating power estimation. The control module can pre-store the resistance value or resistance-temperature coefficient of the wound coil 2 at different temperatures, and calculate the real-time resistance using a resistance-temperature model (e.g., a linear model or a polynomial model) based on the real-time measured temperature of the wound coil 2 (or the temperature estimated by other means) and the current signal, thereby calculating the copper loss. Alternatively, the control module can use online resistance measurement technology to obtain the resistance value of the wound coil 2 in real time by injecting a small signal or using the transformer's own DC resistance measurement function, and then calculate the copper loss by combining it with the current signal.

[0043] Meanwhile, to comprehensively consider the heat sources of the transformer, this application incorporates a loss model of the transformer core 1 to calculate iron losses. Iron losses are the losses generated by the transformer core 1 under the action of an alternating magnetic field, mainly including eddy current losses and hysteresis losses, which are also converted into heat.

[0044] By combining the loss model of transformer core 1 to calculate iron loss, the heat source of the transformer can be fully considered, further improving the accuracy of heat power estimation. The control module can pre-establish the loss model of transformer core 1. This model can be based on empirical formulas (such as the Steinmetz formula) or obtained by fitting experimental data. Input parameters include operating frequency, magnetic flux density (which can be derived from voltage signal), etc., thereby calculating real-time iron loss.

[0045] Another implementation is that the control module can store a pre-calibrated iron loss lookup table, which records the iron loss values ​​under different voltage, frequency and temperature conditions. During operation, the control module looks up the corresponding iron loss value in the table based on the real-time collected voltage and frequency signals.

[0046] Finally, the sum of copper losses and iron losses is taken as the real-time heating power. This step involves summing the copper losses of the two main heat sources of the transformer, the winding coil 2, and the iron losses of the transformer core 1, to obtain the overall real-time heating power of the transformer. This comprehensive approach ensures the comprehensiveness and accuracy of the heating power estimation, providing a reliable basis for subsequent heat dissipation control.

[0047] The control module simply adds the calculated copper loss and iron loss values ​​to obtain the total real-time heating power. Alternatively, it can weight the calculation results of copper loss and iron loss to reflect the relative importance of different losses to the total heating power.

[0048] The above technical solution can significantly improve the accuracy and response efficiency of heat dissipation control. Specifically, by acquiring the current and voltage signals of the wound coil 2 in real time and calculating the active power, a direct and real-time electrical parameter basis is provided for estimating the heat dissipation power.

[0049] Based on this, the copper loss is accurately calculated by combining the resistance-temperature model of the wound coil 2, which effectively solves the problem of inaccurate copper loss estimation caused by the dynamic change of the resistance of the wound coil 2 with temperature, making the assessment of the heating of the wound coil 2 more realistic.

[0050] Meanwhile, by combining the loss model of transformer core 1 to calculate iron loss, the heat contribution of transformer core 1 is fully considered, avoiding the underestimation of heat generation power caused by ignoring iron loss. Finally, the sum of copper loss and iron loss is used as the real-time heat generation power, which integrates all the main heat sources inside the transformer and provides a highly accurate and real-time heat generation power input. This overcomes the problem of inaccurate estimation caused by not considering the resistance and temperature characteristics of the winding coil 2 and the loss of transformer core 1 in traditional methods. It provides a more reliable basis for the basic cooling control signal output by the control module, so that the semiconductor cooling chip 5 and piezoelectric fan 4 can respond more accurately to the actual heat load of the transformer, effectively avoiding temperature overshoot or insufficient heat dissipation, and ensuring the stable operation and long life of the high-frequency transformer under various operating conditions.

[0051] Among them, the resistance-temperature model of the wound coil 2 and the loss model of the transformer core 1 are used to calculate copper loss and iron loss, thereby obtaining real-time heating power. However, in the process of implementation, the model lacks a specific implementation method, which may lead to calculation errors and affect the accuracy of heat dissipation control.

[0052] To address this, a high-frequency transformer with automatic heat dissipation function was proposed. In the resistance-temperature model of the winding coil 2, the real-time resistance of the winding coil 2 is calculated based on the resistance value, resistance temperature coefficient and temperature difference at the reference temperature. The loss model of the transformer core 1 is realized through a pre-calibrated voltage, frequency and iron loss mapping table.

[0053] In the resistance-temperature model of winding coil 2, the real-time resistance calculation formula for winding coil 2 is: in, Reference temperature The resistance value below, The temperature coefficient of resistance is given by the following method: The reference resistance is obtained by measuring the DC resistance of the winding coil 2 using a DC bridge or digital multimeter when the transformer is not energized and the ambient temperature is known (e.g., 25 degrees Celsius). This ambient temperature is the reference temperature. The temperature coefficient of resistance can be directly found in the standard physical parameters of the winding coil material (such as copper). The typical value for copper is 0.00393 per degree Celsius. Alternatively, the resistance can be obtained through experimental calibration: measure the DC resistance of the wound coil 2 at two different known temperatures, then subtract the resistance value at the previous temperature from the resistance value at the later temperature, divide by the resistance value at the reference temperature, and finally divide by the temperature difference to calculate the resistance. These parameters are pre-stored in the control module's memory for real-time calculation.

[0054] The loss model of transformer core 1 is realized through a pre-calibrated voltage, frequency and iron loss mapping table.

[0055] The basis and method for pre-calibrating the transformer core 1 loss model: The losses of transformer core 1 mainly include hysteresis loss and eddy current loss, the magnitude of which is closely related to the transformer's operating voltage and frequency. To obtain accurate iron loss values, this invention uses a pre-calibrated voltage, frequency, and iron loss mapping table to realize the transformer core 1 loss model. The calibration process of this mapping table is as follows: During the testing phase before the transformer leaves the factory, the transformer under test is placed in a constant temperature environment, and different combinations of AC voltage and frequency are applied, and the corresponding input active power is measured.

[0056] Since the transformer was under no-load or very light-load conditions during the test, the copper loss of winding coil 2 was negligible. Therefore, the measured active power could be approximated as the loss of transformer core 1. Each set of voltage and frequency values ​​was recorded in correspondence with the measured iron loss values ​​to form a two-dimensional mapping table. This mapping table covers the actual operating voltage and frequency range of the transformer, and intermediate values ​​between discrete points can be obtained through linear interpolation.

[0057] After calibration, the mapping table is stored in the memory of the control module. During actual operation, the control module collects the voltage and frequency of the winding coil 2 in real time, and quickly obtains the estimated value of iron loss under the current working condition by looking up the table and interpolation. This method is based on experimental data rather than theoretical models, which eliminates the uncertainty in theoretical derivation and the influence of manufacturing tolerances, and ensures the accuracy of iron loss calculation and engineering feasibility.

[0058] Specifically, the formula for calculating the real-time resistance of winding coil 2 is as follows: This formula is based on the physical property that the resistance of a conductor changes linearly with temperature, where, Indicates at a specific reference temperature The resistance value of the lower winding coil 2. This is the reference resistance value. The resistance of the wound coil 2 can be measured in a known and stable temperature environment (e.g., a constant temperature chamber at 25°C) using a high-precision resistance measuring device (such as a digital bridge or a precision multimeter).

[0059] Alternatively, after the transformer is first energized or has been running stably under no-load for a period of time, its DC resistance can be measured in conjunction with the ambient temperature. Determine . The temperature coefficient of resistance (TCR) is an inherent physical parameter characterizing the degree to which the resistance of the material of the winding coil 2 (such as copper wire) changes with temperature. This coefficient can be obtained by consulting the standard physical parameter handbook for the material of the winding coil 2. Alternatively, it can be calibrated experimentally by measuring the resistance of the winding coil 2 at two different known temperatures and then solving a system of simultaneous equations. value.

[0060] The loss model for transformer core 1 is implemented using a pre-calibrated voltage, frequency, and iron loss mapping table. This mapping table is an empirical data structure used to store the loss values ​​of transformer core 1 under different operating conditions. The establishment of this mapping table is usually completed during the transformer design or manufacturing stage.

[0061] The specific method includes: applying a series of AC voltages with different amplitudes and frequencies to the transformer core 1 in a laboratory environment using a dedicated testing device, and accurately measuring the transformer core 1 loss under each voltage-frequency combination (e.g., using a power analyzer or thermal balance method). These measurement data are compiled into a multidimensional lookup table, where the input parameters are voltage and frequency, and the output parameter is the corresponding iron loss value.

[0062] In actual operation, the control module collects the transformer's operating voltage and frequency in real time, and then, by consulting the pre-calibrated mapping table and combining it with interpolation algorithms (such as linear interpolation, bilinear interpolation, etc.), quickly and accurately obtains the transformer core 1 loss value under the current operating state.

[0063] Another approach is to use a mapping table that is a polynomial or nonlinear function fitted to experimental data. By substituting voltage and frequency as independent variables into the function, the iron loss value can be calculated directly, thus avoiding the complex lookup and interpolation process.

[0064] Specifically, by introducing a real-time resistance calculation formula for the wound coil 2, this method can accurately reflect the change in resistance of the wound coil 2 with temperature based on the resistance value and resistance temperature coefficient at the reference temperature, thereby making the calculation of copper loss closer to the actual working conditions.

[0065] Meanwhile, the loss model of transformer core 1 is realized through a pre-calibrated voltage, frequency and iron loss mapping table, which avoids the simplification error that may exist in traditional theoretical models. It directly uses experimental data to reflect the real loss of transformer core 1 at different operating points, ensuring the accuracy and reliability of iron loss calculation. Together, they constitute an accurate estimate of real-time heat generation power, providing a solid foundation for subsequent heat dissipation control.

[0066] Given that the accuracy of real-time heat generation power is crucial for the effective operation of the heat dissipation system, this significantly improves the precision and response speed of heat dissipation control, effectively avoiding problems such as temperature overshoot or insufficient cooling caused by inaccurate heat generation power estimation, thereby ensuring the stability and reliability of high-frequency transformers under complex operating conditions.

[0067] In some embodiments described above in this application, the time rate of change of dynamic resistance is calculated to generate a feedforward correction signal. However, in its implementation, how to accurately calculate the dynamic resistance and its time rate of change is not described in detail, which may lead to calculation errors or delays, affecting the accuracy and response speed of feedforward control.

[0068] In this regard, this application further proposes that when calculating the time-varying rate of change of dynamic resistance, the control module first considers the active power. and the effective value of current Calculate dynamic resistance The formula for its calculation is: To deduct the fixed inductive reactance component of the wound coil 2, the control module employs the following method: real-time acquisition of the effective voltage value across the winding coil 2. and the effective value of current Simultaneously measure the phase difference between voltage and current. The total impedance modulus of winding coil 2 is: Its resistive components: Due to active power Therefore, the resistive component can also be expressed as This is precisely the dynamic resistor in this application. The calculation formula.

[0069] Based on this, to further eliminate the residual effects that the inductive component may have (such as nonlinear inductive reactance caused by harmonics or transformer core 1 saturation), the control module will directly use the resistance component calculated above. As a dynamic resistor, it does not contain any inductive reactance contribution, thus ensuring the accuracy of subsequent copper loss calculations and temperature extrapolation.

[0070] Subsequently, the dynamic resistance is differentiated over time to obtain the rate of change of resistance. Finally, based on the temperature coefficient of resistance and the reference resistance value, the rate of change of temperature of the wound coil 2 is calculated. Finally, it is converted into the temperature change rate of the wound coil 2 according to the following formula. : in, Reference temperature The resistance value below, is the temperature coefficient of resistance.

[0071] Specifically, the control module calculates the dynamic resistance. In this case, various methods can be used. For example, the control module can acquire the voltage and current signals of the wound coil 2 in real time and sample these signals through a digital signal processor (DSP) or an analog-to-digital converter (ADC) in a microcontroller.

[0072] Subsequently, the active power is calculated using a built-in algorithm. and the effective value of current Then, the dynamic resistance is calculated according to the above formula. Another approach is to integrate a dedicated power measurement integrated circuit (IC) into the control module. This IC can directly output the active power and the effective value of the current, thereby simplifying the calculation process of dynamic resistance and improving real-time performance.

[0073] In the dynamic resistance Perform time differentiation to obtain the rate of change of resistance. At this time, the control module can use the numerical differentiation method, or the simple backward difference method as shown in the above formula, by comparing the current time step. resistance value Compared with the previous sampling time resistance value This is used to approximate the rate of change of resistance.

[0074] To improve computational accuracy and noise immunity, the central difference method can be used. Alternatively, in more complex implementations, a smoother and more accurate rate of change of resistance can be obtained by least-squares fitting of the resistance values ​​over a period of time and then differentiating the fitted function. In addition, digital filters, such as Kalman filters, can be introduced to preprocess the resistance data to reduce the impact of measurement noise on the differentiation results.

[0075] Rate of change of resistance Converted to the rate of temperature change of the wound coil 2 Based on the linear physical relationship between conductor resistance and temperature, the control module pre-stores the temperature coefficient of resistance of the winding coil 2 material. and at a certain reference temperature The resistance value below These parameters are usually obtained through experimental calibration during the transformer design or manufacturing stage and are stored in the memory of the control module.

[0076] The control module uses these preset constants to accurately convert the resistance change rate into the temperature change rate of the wound coil 2 through simple multiplication and division operations. This conversion ensures that the obtained temperature change rate signal can directly and accurately reflect the actual thermal state change trend inside the wound coil 2.

[0077] Through the above technical solution, this application can accurately and in real time obtain the dynamic resistance and its rate of change over time of the wound coil 2, and convert it into the rate of change of temperature of the wound coil 2. This method of directly calculating the dynamic resistance using electrical parameters avoids the measurement delay and error caused by thermal inertia or installation position of traditional temperature sensors, and ensures the real-time performance and accuracy of resistance calculation.

[0078] By performing time differentiation on the dynamic resistance, the control module can capture the instantaneous trend of resistance change, rather than relying solely on the steady-state average value. This allows for earlier detection of temperature change trends and rates, providing timely and reliable data for feedforward control of the heat dissipation system.

[0079] By accurately converting the resistance change rate into the temperature change rate, the control module can directly obtain the actual temperature change rate inside the winding coil 2. This is crucial for achieving proactive and predictive thermal management. Based on this accurate temperature change rate signal, the control module can generate a more accurate and responsive feedforward correction signal in combination with real-time heating power. This effectively suppresses temperature fluctuations in the winding coil 2, avoids temperature overshoot, and significantly improves the thermal management performance and operational reliability of the high-frequency transformer.

[0080] In some of the above-mentioned solutions in this application, the time change rate of dynamic resistance is calculated to generate a feedforward correction signal. However, in its implementation, the inductance component of the wound coil 2 will cause the dynamic resistance calculation to be inaccurate, because the inductance component will introduce reactive power interference to the measurement of active power, thereby affecting the accuracy of the resistance value calculation, and thus causing deviation in the temperature change rate calculation.

[0081] In this regard, this application further proposes that before calculating the dynamic resistance, the control module first deducts the fixed inductive reactance component of the wound coil 2 to eliminate the influence of inductance on the resistance calculation.

[0082] This control module is the intelligent control core of a high-frequency transformer with automatic heat dissipation function. It is responsible for executing complex algorithms and control logic. It can be implemented in various hardware forms. It can be a high-performance microcontroller (MCU) that integrates a central processing unit, memory and various peripheral interfaces, and can efficiently process real-time data and execute control instructions. Alternatively, it can be a digital signal processor (DSP) that is good at fast digital signal processing and complex mathematical operations, and is particularly suitable for accurate calculation of power and resistance. Alternatively, a Field Programmable Gate Array (FPGA) can be used, programmed via a hardware description language to implement highly parallel and customized logic functions to meet the needs of high-speed real-time control. Subtracting the fixed inductive reactance component of the wound coil 2 aims to accurately separate the pure resistive component from the measured total impedance, thereby eliminating the interference of the inherent inductance of the wound coil 2 on the resistance calculation.

[0083] In practical implementation, the control module can compensate through software algorithms. For example, after acquiring the voltage and current signals of the wound coil 2, it can use the fixed inductive reactance value of the wound coil 2, which is pre-calibrated or obtained through theoretical calculation, to perform vector subtraction on the complex impedance plane to obtain the pure resistance value. Alternatively, in the power calculation stage, it can accurately measure the active power and reactive power, and use the reactive power component to correct the deviation between the apparent power and active power caused by the inductive reactance, thereby more accurately calculating the resistance value.

[0084] In addition, a specific hardware filtering circuit can be designed at the signal acquisition front end. Through series or parallel passive or active compensation networks, the inductive reactance effect can be partially canceled before the signal enters the control module, so that the signal received by the control module is closer to the response of a purely resistive load.

[0085] The core purpose of this technique, which eliminates the influence of inductance on resistance calculation, is to ensure the accuracy and reliability of dynamic resistance calculation. By accurately subtracting the fixed inductive reactance component of the wound coil 2, the control module can avoid interference from reactive power introduced by inductance on active power measurement, thereby ensuring the accuracy and reliability of the dynamic resistance calculated based on active power and the effective value of current. It can accurately reflect the ohmic resistance value of the wound coil 2. This is crucial for the subsequent accurate calculation of the temperature change rate of the wound coil 2, because only an accurate resistance value can guarantee the accuracy of temperature calculation, thereby providing a reliable input for the precise control of the heat dissipation system.

[0086] Through the above technical solution, this application effectively solves the problem of interference of the inductance component of the wound coil 2 on the accuracy of dynamic resistance calculation. Specifically, before calculating the dynamic resistance, the control module can accurately separate the pure resistance component of the wound coil 2 from the total impedance by subtracting the fixed inductance component of the wound coil 2, thereby avoiding the influence of reactive power introduced by the inductance on the active power measurement.

[0087] This makes dynamic resistance The calculation results are more accurate, thus ensuring the accuracy of the temperature change rate of the wound coil 2 calculated based on the dynamic resistance time change rate. Given this precision, the control module is able to generate more reliable and responsive feedforward correction signals for precisely adjusting the driving voltage of the semiconductor cooling chip 5 and the driving voltage amplitude of the piezoelectric fan 4.

[0088] This improved precision ensures that the heat dissipation system can respond to the temperature change trend of the winding coil 2 more promptly and accurately, effectively avoiding temperature overshoot and controlling the temperature of the winding coil 2 within a narrower range, thereby significantly improving the operational stability and reliability of the high-frequency transformer under conditions of sudden load changes or high power density operation.

[0089] In some of the solutions mentioned above in this application, a feedforward correction signal is proposed to respond quickly to temperature changes. However, in its implementation, without a specific generation method, the control signal may be inaccurate, affecting the coordination of heat dissipation efficiency and response speed.

[0090] In response, this application further proposes a method for generating the feedforward correction signal. The method for generating the feedforward correction signal is as follows: the temperature change rate of the winding coil 2 is multiplied by a preset first gain coefficient and superimposed on the driving voltage of the semiconductor cooling chip 5 in the basic cooling control signal. At the same time, the temperature change rate of the winding coil 2 is multiplied by a preset second gain coefficient and superimposed on the driving voltage amplitude of the piezoelectric fan 4 in the basic cooling control signal.

[0091] Specifically, the temperature change rate of the wound coil 2 is obtained by converting the time-varying rate of dynamic resistance. Multiply by the preset first gain factor According to the formula: This is superimposed on the driving voltage of the semiconductor cooling chip 5 in the basic cooling control signal. At the same time, the temperature change rate of the wound coil 2 is multiplied by a preset second gain coefficient. According to the formula: The gain coefficient is superimposed on the basic cooling control signal to control the driving voltage amplitude of the piezoelectric fan 4. It should be noted that the first gain coefficient and the second gain coefficient are both determined experimentally, with typical values ​​ranging from 0.01 volts to 0.5 volts per second per degree Celsius.

[0092] The temperature change rate of the wound coil 2 is obtained by converting the time-varying rate of the dynamic resistance. This refers to monitoring the change in the dynamic resistance of the winding coil 2 over time and converting it into the instantaneous rate of change of the internal temperature of the winding coil 2.

[0093] The change in dynamic resistance directly reflects the real-time temperature state of the wound coil 2, while its rate of change over time provides information on the trend and speed of temperature change. This rate of change of temperature of the wound coil 2 can be obtained in a variety of ways.

[0094] For example, by continuously sampling the current and voltage signals of the wound coil 2, the real-time active power and the effective value of the current can be calculated, and then the dynamic resistance can be obtained. Then, by using digital signal processing methods such as differential or filtering, the time change rate of the dynamic resistance can be calculated, and combined with the resistance temperature coefficient of the wound coil 2 and the reference resistance value, it can be converted into the temperature change rate of the wound coil 2.

[0095] Another approach is to use a dedicated resistance-temperature conversion module. This module has a built-in resistance-temperature model of the wound coil 2, which can receive dynamic resistance values ​​in real time and directly output the corresponding temperature change rate signal, thereby simplifying the calculation burden of the control module.

[0096] Preset first gain coefficient It is a pre-set scaling factor used to adjust the influence of the temperature change rate of the wound coil 2 on the driving voltage of the thermoelectric cooler 5. The setting of this coefficient is directly related to the sensitivity and intensity of the thermoelectric cooler 5's response to temperature changes. This can be obtained through experimental calibration, that is, by testing the dynamic response of the transformer's heat dissipation system under different loads and ambient temperatures, and adjusting accordingly. To achieve optimal temperature control and stability.

[0097] In addition, simulation modeling and optimization algorithms can also be used to determine For example, establishing a transformer thermal model and a heat dissipation system control model, and using PID or other advanced control algorithms to control the heat dissipation system. Offline optimization is performed to ensure that the system can respond quickly and stably to temperature changes under various operating conditions.

[0098] The driving voltage used in the basic cooling control signal to control the semiconductor cooling chip 5 It is the basic drive voltage received by the semiconductor cooling chip 5 without feedforward correction. This voltage is typically generated by the basic cooling control logic based on real-time heat generation power calculations, which is designed to provide the reference cooling capacity required to maintain the normal operating temperature of the transformer.

[0099] This can be obtained by mapping real-time heating power to a preset cooling power curve or lookup table. When the real-time heating power increases, This will also be increased accordingly to provide greater cooling capacity, or it can be generated through a closed-loop feedback controller (such as a PID controller). The controller uses the average temperature of the wound coil 2 or the temperature of the transformer core 1 as feedback and the target temperature as the set value to output a stable base drive voltage.

[0100] Preset second gain coefficient It is another pre-set proportional factor used to adjust the degree of influence of the temperature change rate of the wound coil 2 on the driving voltage amplitude of the piezoelectric fan 4. The setting of this coefficient determines the response speed of the piezoelectric fan 4 to temperature changes and the air volume adjustment range.

[0101] Similar to , Alternatively, it can be determined through experimental calibration, observing the airflow regulation effect of the piezoelectric fan 4 during actual operation, and optimizing it according to heat dissipation requirements and noise limits. Likewise, simulation analysis can be used, combined with the characteristic curves of the piezoelectric fan 4 (such as the relationship between drive voltage amplitude and airflow), to optimize its performance. Optimizations were made to ensure that the piezoelectric fan 4 could work in conjunction with the semiconductor cooling chip 5 to achieve efficient heat dissipation.

[0102] The basic refrigeration control signal is used to control the amplitude of the drive voltage of the piezoelectric fan 4. It is the basic drive voltage amplitude received by the piezoelectric fan 4 without feedforward correction. This amplitude is usually generated by the basic cooling control logic based on the overall heat dissipation requirements of the transformer, and is designed to provide the reference airflow required to maintain the normal operating temperature of the transformer. It can be obtained by mapping real-time heating power with preset air volume demand curves or lookup tables.

[0103] For example, when the heating power is high, This will also be increased accordingly to provide a larger air volume, or it can be generated through an independent closed-loop feedback controller. The controller uses the air temperature or heat sink temperature inside the transformer as feedback and the target heat dissipation effect as the set value, thereby outputting a stable base drive voltage amplitude.

[0104] Through the above technical solution, this application can effectively solve the problems of inaccurate control signals, poor coordination of heat dissipation efficiency, and slow response speed caused by the lack of a specific feedforward correction signal generation method in traditional heat dissipation control.

[0105] Specifically, the temperature change rate of the wound coil 2, obtained by converting the time change rate of the dynamic resistance, is multiplied by a preset first gain coefficient. This voltage is then superimposed on the basic driving voltage of the thermoelectric cooler 5, enabling precise and forward-looking adjustment of the cooling intensity of the thermoelectric cooler 5.

[0106] This method directly utilizes the temperature trend information contained in the resistance change of the winding coil 2, effectively avoiding the measurement delay problem of traditional temperature sensors, enabling the semiconductor cooling chip 5 to respond to the temperature change inside the winding coil 2 earlier and more accurately, thereby suppressing temperature overshoot.

[0107] At the same time, the temperature change rate of the same wound coil 2 is multiplied by a preset second gain coefficient. This is then superimposed on the base drive voltage amplitude of the piezoelectric fan 4, ensuring that the airflow adjustment of the piezoelectric fan 4 is synchronized with the cooling intensity of the semiconductor cooling chip 5, enhancing the coordination of the entire heat dissipation system. This is achieved by adjusting the gain coefficient. and With preset and optimized features, the system can flexibly adjust its response characteristics according to actual operating conditions, thereby achieving fast, stable and efficient heat dissipation control when the load changes suddenly or the ambient temperature changes, which significantly improves the operational reliability and performance stability of the high-frequency transformer.

[0108] In some embodiments of this application, a control module is proposed to dynamically adjust the heat dissipation device based on real-time heat generation power and dynamic resistance change rate. However, in the implementation process, the performance degradation or failure of the heat dissipation system itself (such as the semiconductor cooling chip 5 or the piezoelectric fan 4) may lead to a decrease in the actual heat dissipation capacity. The existing solution lacks a direct detection mechanism for the operating efficiency of the heat dissipation system, and cannot identify heat dissipation abnormalities and issue warnings in a timely manner, thereby affecting the temperature control stability of the transformer.

[0109] In response, this application further proposes a method for calculating the efficiency deviation of the heat dissipation system as follows: based on the real-time heat dissipation power, the actual heat dissipation power of the current semiconductor cooling chip 5 and the piezoelectric fan 4, and the heat capacity of the transformer body, the expected temperature rise rate is calculated, and then the actual temperature rise rate obtained by the dynamic resistance time change rate is subtracted from the expected temperature rise rate to obtain the efficiency deviation; when the efficiency deviation is greater than a preset threshold, the efficiency of the heat dissipation system is determined to be abnormal.

[0110] Specifically, based on real-time heating power The actual heat dissipation power of the current semiconductor cooling chip 5 and piezoelectric fan 4 and the heat capacity of the transformer body ; Among them, the actual heat dissipation power of the current semiconductor cooling chip 5 and piezoelectric fan 4 The following method is used for estimation: Actual heat dissipation power of semiconductor cooling chip 5 By monitoring its operating voltage and operating current Combined with the pre-calibrated cooling efficiency calculate: Cooling efficiency It can be calibrated experimentally before leaving the factory: under the condition that the hot end of the semiconductor cooler 5 is kept at a constant temperature (e.g., 25°C), different... , Calculate the actual cooling capacity under the combined load: The typical value range is 0.6 to 0.8, and the specific value is determined based on the actual calibration results.

[0111] Heat dissipation power of piezoelectric fan 4 Through its driving voltage amplitude Estimate the air volume, then calculate based on the heat transfer coefficient at the hot end: First, air volume The relationship with the driving voltage amplitude can be calibrated as linear: Where the coefficient The airflow was obtained by fitting the data after measuring the airflow at different voltages with an anemometer. Then, the inlet air temperature at the hot end of the semiconductor cooling chip 5 was measured. and air outlet temperature Calculate the temperature difference: The heat dissipation power of piezoelectric fan 4 is: in The specific heat capacity of air, This refers to air density.

[0112] If the temperature difference between the inlet and outlet air cannot be measured, an empirical formula can be used as an approximation: in, The coefficient is calibrated through a thermal equilibrium experiment.

[0113] Total actual heat dissipation power = .

[0114] According to the formula: Calculate the expected temperature rise rate, and then use the actual temperature rise rate obtained from the dynamic resistance change rate over time. Subtract from the expected rate of temperature rise, according to the formula: get ;when If the value exceeds a preset threshold, the cooling system efficiency is considered abnormal.

[0115] Among them, real-time heating power This refers to the total heat generated during the operation of a high-frequency transformer by factors such as the copper loss of the winding coil 2 and the iron loss of the transformer core 1.

[0116] This power is a quantitative representation of the heat source inside the transformer and is the basis for assessing heat dissipation requirements. It can be obtained by real-time acquisition of the current and voltage signals of the wound coil 2, combined with a transformer loss model for calculation, to determine the actual heat dissipation power of the current semiconductor cooling chip 5 and piezoelectric fan 4. This refers to the amount of heat actually removed from the transformer body by the heat dissipation system consisting of the semiconductor cooling chip 5 and the piezoelectric fan 4 under the current operating conditions.

[0117] The actual heat dissipation power of the semiconductor cooling chip 5 can be estimated by monitoring its operating current and voltage, and by referring to a table based on the cooling performance curve (i.e., the current-cooling capacity relationship) provided by the manufacturer. The actual heat dissipation power of the piezoelectric fan 4 can be estimated by the calibration relationship between its driving voltage amplitude and airflow, combined with the hot-end heat transfer coefficient model. The above estimation methods are all conventional techniques in this field and will not be elaborated further here.

[0118] This power reflects the actual working capacity of the heat dissipation system. It can be obtained by monitoring parameters such as the driving current and voltage of the semiconductor cooling chip 5 and the driving voltage amplitude of the piezoelectric fan 4, and by combining them with the pre-calibrated or real-time estimated performance curve of the heat dissipation components, or by measuring parameters such as the temperature difference and flow rate at the inlet and outlet of the heat dissipation components.

[0119] The heat capacity of the transformer body Heat capacity refers to the amount of heat absorbed or released by the transformer body when the temperature increases by one degree Celsius. It is an important physical parameter characterizing the heat storage capacity of a transformer, and it is related to factors such as the type and quality of the materials used in the transformer, as well as its structure.

[0120] This parameter can be obtained by summing the heat capacities of the materials of each component of the transformer, or by calibrating it experimentally to predict the temperature rise rate. It is based on the principle of thermodynamic equilibrium, according to the real-time heat generation power inside the transformer. Actual heat dissipation power of the heat dissipation system and the heat capacity of the transformer body The calculated theoretical rate of temperature change represents the expected trend of transformer body temperature change under current heating and cooling conditions. The actual temperature rise rate is obtained from the dynamic resistance-time change rate. The actual rate of temperature change of the winding coil 2 is obtained by real-time monitoring of the dynamic resistance change of the winding coil 2 and converting it into a value.

[0121] This method can quickly and accurately reflect the instantaneous temperature change inside the wound coil 2, avoiding the measurement delay problem of traditional temperature sensors. It is the difference between the expected rate of temperature rise and the actual rate of temperature rise. This difference quantifies the degree of deviation between the actual performance of the heat dissipation system and the theoretical expectation, and is a key indicator for evaluating the efficiency of the heat dissipation system.

[0122] The preset threshold is a pre-set value used as a reference standard to determine whether the efficiency of the heat dissipation system is abnormal. This threshold can be set according to factors such as transformer design requirements, operating environment, safety margin, and experimental data to ensure that an early warning can be issued in time when the heat dissipation efficiency drops slightly.

[0123] Determining abnormal efficiency of the heat dissipation system refers to When the preset threshold is exceeded, the control module determines that the working efficiency of the heat dissipation system has deviated from the normal range, and there may be performance degradation or potential failure. This determination will trigger corresponding warning or protection measures.

[0124] By introducing an efficiency deviation calculation mechanism for the heat dissipation system, this application can effectively solve the problem that it is difficult to detect the performance degradation or failure of the heat dissipation device in a timely manner.

[0125] Specifically, by calculating the expected temperature rise rate of the transformer in real time and comparing it with the actual temperature rise rate obtained from the dynamic resistance time change rate, the deviation between the actual working efficiency of the heat dissipation system and the theoretical expectation can be accurately and quickly quantified.

[0126] When the deviation exceeds the preset threshold, the control module can immediately determine that the heat dissipation system efficiency is abnormal and output an early warning signal. This mechanism avoids the response lag problem caused by the thermal conduction delay of traditional temperature sensors and ensures real-time monitoring of the heat dissipation system status.

[0127] Therefore, even in the early stages of performance degradation or failure of the semiconductor cooling chip 5 or piezoelectric fan 4, timely detection and intervention measures can be taken, thereby effectively maintaining the temperature control stability of the high-frequency transformer, avoiding the risk of overheating due to insufficient heat dissipation, and significantly improving the reliability and safety of transformer operation.

[0128] In some of the solutions described above in this application, a piezoelectric fan 4 is proposed for heat dissipation. However, the non-fixed drive frequency of the fan may lead to low efficiency and inaccurate response, because driving at a non-resonant frequency will reduce energy conversion efficiency and increase power consumption. At the same time, the inertial delay introduced by frequency changes will hinder the rapid adjustment of airflow and affect the timely response of the heat dissipation system to sudden load changes.

[0129] In this regard, this application further proposes that the driving frequency of the piezoelectric fan 4 is fixed to the first-order resonant frequency of its fan blades, and the control module continuously changes the amplitude of the fan blades by adjusting the amplitude of the driving voltage, thereby adjusting the air volume.

[0130] Specifically, the piezoelectric fan 4 is a device that uses the piezoelectric effect to drive the fan blades to vibrate and generate airflow. It drives the fan blades to reciprocate through the deformation of the piezoelectric material and has the advantages of small size, low power consumption and no electromagnetic noise.

[0131] The first-order resonant frequency refers to the natural frequency at which the vibration amplitude of the fan blade reaches its maximum under a specific driving frequency. Fixing the driving frequency to the first-order resonant frequency of the fan blade means that when the piezoelectric fan 4 is working, the frequency of its driving signal is set to be consistent with the natural resonant frequency of the fan blade.

[0132] This can be achieved in several ways. For example, after the piezoelectric fan 4 is manufactured or installed, the first-order resonant frequency of its blades can be determined through experimental testing, and this frequency value can be preset and stored in the control module. During actual operation, the control module directly drives the piezoelectric fan 4 at this preset frequency.

[0133] Alternatively, the control module can integrate a resonant frequency tracking algorithm to monitor the current or vibration feedback signal of the piezoelectric fan 4 and adjust the drive frequency in real time so that it is always locked at the maximum vibration response frequency of the fan blades, in order to adapt to the resonant frequency drift caused by environmental changes or component aging.

[0134] Based on this, the control module is responsible for generating and outputting the electrical signal required to drive the piezoelectric fan 4. The frequency of this signal is fixed at the first-order resonant frequency, while its voltage amplitude is adjustable. Adjusting the amplitude of the drive voltage means that the peak voltage or effective voltage of the AC drive signal output by the control module to the piezoelectric fan 4 is variable.

[0135] This can be achieved through pulse width modulation (PWM) technology, whereby the control module effectively controls the average voltage or equivalent voltage amplitude applied to the piezoelectric fan 4 without changing the frequency by changing the duty cycle of the pulse.

[0136] Alternatively, the control module can use a digital-to-analog converter (DAC) combined with a power amplifier to directly generate and output an AC voltage signal with a sine wave or other waveforms with variable amplitude. By changing the output value of the DAC, the amplitude of the driving voltage can be precisely adjusted. Since the degree of deformation of the piezoelectric material (and thus the amplitude of the fan blade) is proportional to the amplitude of the voltage applied to it, the smooth and stepless adjustment of the fan blade amplitude can be achieved by continuously adjusting the amplitude of the driving voltage.

[0137] The amplitude of the fan blades directly determines the amount of air pushed by each vibration. The larger the amplitude, the more air is pushed and the greater the air volume generated. Conversely, the smaller the amplitude, the greater the air volume generated. Therefore, by adjusting the amplitude of the fan blades, the output air volume of the piezoelectric fan can be precisely controlled.

[0138] By using the above technical solution, the driving frequency of the piezoelectric fan 4 is fixed at the first-order resonant frequency of its blades, ensuring that the fan operates at the mechanical resonance point, thereby maximizing energy conversion efficiency, reducing unnecessary power consumption loss, and avoiding instability caused by frequency drift.

[0139] Meanwhile, the control module continuously changes the amplitude of the fan blades by adjusting the amplitude of the drive voltage, which enables precise and rapid adjustment of the air volume. This amplitude adjustment method avoids the inertial delay caused by frequency adjustment, allowing the heat dissipation system to respond promptly to the temperature change trend of the winding coil 2 and effectively suppressing the temperature overshoot phenomenon.

[0140] Overall, this solution, by combining fixed frequency with amplitude adjustment, significantly improves the response speed and control accuracy of the heat dissipation system while ensuring heat dissipation efficiency. This allows the temperature of the winding coil 2 of the high-frequency transformer to be controlled within a narrower range under conditions of sudden load changes or precise control, thereby ensuring the stability and reliability of the transformer's electrical parameters.

[0141] In some of the solutions described above in this application, a control module is proposed to output an early warning signal when the efficiency deviation exceeds a threshold. However, simply outputting an early warning signal may not be sufficient to actively prevent the transformer from overheating or degrading. Further measures are needed to maintain safe operation and heat dissipation efficiency.

[0142] In response, after determining that the efficiency deviation exceeds the threshold, the control module also performs at least one of the following operations: automatically reducing the maximum allowable input power of the transformer; increasing the driving voltage amplitude of the piezoelectric fan 4 to compensate for the performance degradation of the semiconductor cooling chip 5.

[0143] Specifically, when the control module determines that the efficiency deviation of the heat dissipation system exceeds a preset threshold, it can take proactive intervention measures. One such measure is automatically reducing the transformer's maximum allowable input power, which aims to directly alleviate heat dissipation pressure by reducing heat source generation. This operation can be implemented in several ways. For example, the control module can send instructions to the transformer's power supply management unit to adjust its output voltage or current limits, thereby limiting the transformer's maximum input power. In another implementation, if the transformer supplies power to a controllable load, the control module can communicate with the load management system, requesting it to reduce power consumption requirements, thereby indirectly reducing the transformer's load and input power.

[0144] Simultaneously, the control module can also selectively or in combination execute operations to increase the drive voltage amplitude of the piezoelectric fan 4 to compensate for the performance degradation of the thermoelectric cooler 5. This operation aims to compensate for potential performance decline in the thermoelectric cooler 5 by enhancing convection heat dissipation. The control module can achieve the increase in drive voltage amplitude in various ways.

[0145] For example, the effective voltage amplitude applied to the piezoelectric fan 4 can be increased by increasing the duty cycle of the pulse width modulation (PWM) signal, thereby increasing its vibration amplitude and generating a larger air volume. Alternatively, the control module can adjust the gain of a variable gain amplifier located in the drive signal path to directly increase the voltage signal amplitude applied to the piezoelectric fan 4.

[0146] With the above technical solution, when the efficiency deviation of the heat dissipation system exceeds the preset threshold, the control module will no longer just issue a warning, but will be able to take proactive intervention measures, thereby effectively preventing the transformer from overheating or further degrading in performance.

[0147] Specifically, automatically reducing the maximum allowable input power of the transformer can reduce heat generation at the source, preventing temperature runaway due to continuous high power operation when heat dissipation capacity is insufficient, thereby protecting the internal components of the transformer and extending its service life.

[0148] At the same time, increasing the driving voltage amplitude of the piezoelectric fan 4 can enhance the convective heat transfer capability of the heat dissipation system. Even if the cooling efficiency of the semiconductor cooling chip 5 decreases due to aging or other reasons, its performance degradation can be compensated by increasing the air volume, maintaining the overall heat dissipation efficiency and ensuring that the system can still operate stably under abnormal conditions.

[0149] This proactive intervention mechanism, which combines limiting the heat generation power with enhancing the heat dissipation capacity, enables high-frequency transformers with automatic heat dissipation functions to achieve a higher level of self-adaptation and fault tolerance when faced with abnormalities in the heat dissipation system, significantly improving the reliability and safety of the system.

[0150] During transformer operation, the control module acquires the current and voltage signals of the wound coil 2 in real time. Unlike traditional schemes that rely on feedback from hysteresis temperature sensors, this system directly utilizes these electrical signals for rapid response.

[0151] The control module first calculates the active power based on the acquired current and voltage signals. Next, it calculates copper losses using dynamic resistance, while a resistance-temperature model is used for temperature estimation, and iron losses are calculated in conjunction with the loss model of transformer core 1. For example, the real-time resistance of the winding coil 2... Through the formula: The sum of copper loss and iron loss is the real-time heat generation power of the transformer. Based on this real-time heat generation power, the control module outputs a basic cooling control signal to initially drive the semiconductor cooling chip 5 and the piezoelectric fan 4 for heat dissipation.

[0152] It should be noted that the "resistance-temperature model" and "dynamic resistance" in this application have different uses: the dynamic resistance is directly calculated by electrical measurement and used for copper loss calculation to obtain real-time heating power, without relying on any prior temperature knowledge; the resistance-temperature model is used to deduce the real-time temperature or temperature change rate of the winding coil 2 from the dynamic resistance to generate a feedforward correction signal. The two have a clear division of labor, avoiding the temperature assumptions required when using the resistance-temperature model for copper loss calculation in traditional schemes.

[0153] To further improve response speed and avoid temperature overshoot, the control module also calculates the time-varying rate of dynamic resistance using the same current and voltage signals. Specifically, the control module first calculates the time-varying rate of dynamic resistance based on the active power... and the effective value of current Calculate dynamic resistance Calculation formula: Before calculating the dynamic resistance, the control module subtracts the fixed inductive reactance component of the wound coil 2 to eliminate the influence of inductance on the resistance calculation and ensure the accuracy of the calculation. Then, the dynamic resistance is calculated... By differentiating over time, the rate of change of resistance is obtained: Finally, according to the formula: Convert the rate of change of resistance to the rate of change of temperature of the wound coil 2. The temperature change rate of this wound coil 2 It can reflect the temperature change trend of the wound coil 2 in real time and quickly, and has predictive ability, overcoming the drawback of measurement delay of traditional temperature sensors.

[0154] Based on the temperature change rate of this winding coil 2 The control module generates a feedforward correction signal. For example, when a rapid increase in the temperature of the wound coil 2 is detected (i.e., (If the value is positive and relatively large), the control module will multiply the temperature change rate of the wound coil 2 by a preset first gain coefficient. This signal is then superimposed on the basic cooling control signal that controls the driving voltage of the semiconductor cooling chip 5 to form the final driving voltage. At the same time, the temperature change rate of the wound coil 2 is multiplied by a preset second gain coefficient. This is superimposed on the basic cooling control signal that controls the amplitude of the drive voltage of the piezoelectric fan 4 to form the final drive voltage amplitude: The driving frequency of the piezoelectric fan 4 is fixed to the first resonant frequency of its blades. The amplitude of the blades is continuously changed by adjusting the amplitude of the driving voltage, thereby adjusting the air volume.

[0155] This feedforward correction mechanism enables the heat dissipation system to adjust the cooling and airflow in advance based on the temperature change trend before the actual temperature rises significantly, thereby achieving precise and rapid control of the temperature of the wound coil 2 and effectively suppressing temperature overshoot.

[0156] Compared with the DC resistance-based temperature estimation scheme used only for overheat protection, this scheme utilizes the resistance change rate for continuous dynamic adjustment, which significantly improves the precision and response speed of the control.

[0157] In addition, to ensure the long-term reliability of the heat dissipation system, the control module also calculates the efficiency deviation of the heat dissipation system, based on the real-time heat generation power. The actual heat dissipation power of the current semiconductor cooling chip 5 and piezoelectric fan 4 and the heat capacity of the transformer body According to the formula: Calculate the expected temperature rise rate, and then use the actual temperature rise rate obtained from the dynamic resistance change rate over time. Subtracting from the expected rate of temperature rise, we get: when When the efficiency exceeds a preset threshold, the control module determines that the cooling system efficiency is abnormal and outputs a warning signal. For example, if... A persistently high level may indicate a performance degradation of the thermoelectric cooler 5 or a malfunction of the piezoelectric fan 4.

[0158] In this situation, the control module can perform further operations, such as automatically reducing the maximum allowable input power of the transformer, or increasing the driving voltage amplitude of the piezoelectric fan 4 to compensate for the performance degradation of the semiconductor cooling chip 5, thereby taking preventive measures before a fault occurs and ensuring the safe and stable operation of the transformer. This efficiency deviation monitoring and early warning mechanism provides additional protection for the long-term reliability of high-frequency transformers.

[0159] The terms "front," "back," "left," "right," "top," and "bottom" all refer to the figures in the accompanying drawings. Figure 1 Based on the perspective of the observer, the side of the device facing the observer is defined as the front, the left side of the observer is defined as the left, and so on.

[0160] In the description of this invention, it should be understood that the terms "center", "longitudinal", "lateral", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limiting the scope of protection of this invention.

[0161] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A high-frequency transformer with automatic heat dissipation function, characterized in that, include: Transformer core, A wound coil, the wound coil being wound on the transformer core; A fixed frame, which covers the outside of the transformer core and the winding coil; A heat transfer plate, which is mounted on the fixed frame; A semiconductor refrigeration chip, wherein the semiconductor refrigeration chip is fixedly mounted on the outer wall of a heat transfer plate; A piezoelectric fan, wherein the piezoelectric fan is mounted near the hot end of the semiconductor cooling chip and directly facing the hot end; A control module, which is electrically connected to both the thermoelectric cooler and the piezoelectric fan, is configured to: The system acquires the current and voltage signals of the wound coil in real time, calculates the real-time heat generation power based on the current and voltage signals, and outputs a basic cooling control signal accordingly. It also calculates the time change rate of the dynamic resistance based on the same current and voltage signals, generates a feedforward correction signal, and superimposes it on the basic cooling control signal. Furthermore, it calculates the efficiency deviation of the heat dissipation system based on the real-time heat generation power and the time change rate of the dynamic resistance, and outputs a warning signal when the efficiency deviation exceeds a threshold. When the control module calculates the real-time heating power, it first calculates the active power and the effective value of the current from the current signal and the voltage signal, calculates the dynamic resistance of the winding coil based on the active power and the effective value of the current, then calculates the copper loss using the dynamic resistance, and calculates the iron loss by combining the loss model of the transformer core. The sum of the copper loss and the iron loss is the real-time heating power. When calculating the time-varying rate of change of the dynamic resistance, the control module first considers the active power. and the effective value of current Calculate dynamic resistance Calculation formula: ; The resistance change rate is then obtained by differentiating the dynamic resistance over time. Finally, the temperature change rate of the wound coil is calculated based on the temperature coefficient of resistance and the reference resistance value. The feedforward correction signal is generated by multiplying the temperature change rate of the winding coil by a preset first gain coefficient and superimposing it onto the driving voltage of the semiconductor cooling chip in the basic cooling control signal. At the same time, the temperature change rate of the winding coil is multiplied by a preset second gain coefficient and superimposed on the driving voltage amplitude of the piezoelectric fan in the basic cooling control signal. The efficiency deviation of the heat dissipation system is calculated as follows: the expected temperature rise rate is calculated based on the real-time heat dissipation power, the actual heat dissipation power of the current semiconductor cooling chip and the piezoelectric fan, and the heat capacity of the transformer body. The actual temperature rise rate obtained from the dynamic resistance change rate over time is then subtracted from the expected temperature rise rate to obtain the efficiency deviation; when the efficiency deviation is greater than a preset threshold, the heat dissipation system efficiency is determined to be abnormal.

2. A high-frequency transformer with automatic heat dissipation function according to claim 1, characterized in that: The control module also calculates the real-time temperature or temperature change rate of the wound coil based on the dynamic resistance and resistance-temperature model. The resistance-temperature model is as follows: The real-time resistance of the wound coil is equal to the resistance value at the reference temperature multiplied by one, plus the resistance temperature coefficient multiplied by the temperature difference, where the resistance value at the reference temperature and the resistance temperature coefficient are pre-stored constants. The loss model of the transformer core is realized through a pre-calibrated voltage, frequency and iron loss mapping table.

3. A high-frequency transformer with automatic heat dissipation function according to claim 2, characterized in that: Before calculating the dynamic resistance, the control module first deducts the fixed inductive reactance component of the wound coil to eliminate the influence of inductance on the resistance calculation.

4. A high-frequency transformer with automatic heat dissipation function according to claim 3, characterized in that: The driving frequency of the piezoelectric fan is fixed at the first-order resonant frequency of its blades. The control module continuously changes the amplitude of the blades by adjusting the amplitude of the driving voltage, thereby adjusting the airflow.

5. A high-frequency transformer with automatic heat dissipation function according to claim 4, characterized in that: After determining that the efficiency deviation exceeds the threshold, the control module also performs at least one of the following operations: automatically reducing the maximum allowable input power of the transformer; Increase the driving voltage amplitude of the piezoelectric fan to compensate for the performance degradation of the semiconductor cooling chip.

Images