An intelligent cooling fan system against electromagnetic interference based on the design of a dynamic filter circuit

Abstract

The present invention relates to an anti-electromagnetic interference intelligent cooling fan system based on the design of a dynamic filtering circuit. The system includes: a dual-analysis module for performing dual analysis in the time domain and frequency domain on the rotational speed feedback signal of a double ball bearing fan to obtain a multi-dimensional electromagnetic interference characteristic model, and configuring the filtering network parameters to obtain a target three-stage cascaded dynamic filtering network; a signal purification module for collecting the first fan operation parameters and inputting them into the target three-stage cascaded dynamic filtering network for signal purification to obtain the filtered Hall sensor signal and the filtered power supply signal; a closed-loop control calculation module for performing closed-loop control calculations to obtain a non-linear PWM duty cycle control quantity and a phase distribution parameter; and a phase difference control module for performing phase difference control to obtain a multi-fan collaborative control strategy. The present invention realizes high-precision control and anti-interference ability within a wide rotational speed range, and solves the problem of mutual influence of electromagnetic interference when multiple fans are connected in parallel.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent heat dissipation control, and particularly to an anti-electromagnetic interference intelligent heat dissipation fan system based on the design of a dynamic filter circuit. Background Art

[0002] Heat dissipation fans are widely used as cooling components in servers, communication devices, and industrial control systems. However, these devices usually operate in complex electromagnetic environments, and electromagnetic interference generated by various switching power supplies, frequency conversion devices, and radio frequency devices will seriously affect the stability of the fan control system. Especially when the fan operates within a wide speed range, the speed feedback signal will exhibit severe time-varying non-linear characteristics. Traditional heat dissipation fan control systems usually adopt static filtering schemes, which cannot effectively cope with the changes in electromagnetic interference characteristics under different working conditions, resulting in fan speed fluctuations, increased noise, and shortened service life.

[0003] In high-density computing environments, especially when multiple fans operate in parallel in the data center cabinet heat dissipation system, the electromagnetic interference problem deteriorates further. Since multiple fans operate in the high-speed range simultaneously, their PWM control signals affect each other, forming interference coupling, and traditional control systems cannot achieve collaborative suppression of interference in multi-fan systems. At the same time, the existing fan control systems lack an effective health status assessment mechanism and cannot detect control anomalies in a timely manner and make adaptive adjustments. When the fan runs for a long time, the deviation between the control parameters and the actual requirements accumulates continuously, ultimately resulting in a decrease in heat dissipation efficiency and an increase in energy consumption. Summary of the Invention

[0004] The main object of the present invention is to provide an anti-electromagnetic interference intelligent heat dissipation fan system based on the design of a dynamic filter circuit. The present invention realizes high-precision control and anti-interference ability within a wide speed range, and solves the problem of mutual influence of electromagnetic interference when multiple fans are connected in parallel.

[0005] To achieve the above object, the present invention provides an anti-electromagnetic interference intelligent heat dissipation fan system based on the design of a dynamic filter circuit, including:

[0006] A dual-analysis module for performing dual analysis in the time domain and frequency domain on the speed feedback signal of a double-ball bearing fan to obtain a multi-dimensional electromagnetic interference characteristic model, and configuring the filter network parameters of an initial three-stage cascaded dynamic filter network to obtain a target three-stage cascaded dynamic filter network;

[0007] A signal purification module for collecting the first fan operation parameter of the double-ball bearing fan and inputting it into the target three-stage cascaded dynamic filter network for signal purification to obtain a filtered Hall sensor signal and a filtered power supply signal;

[0008] A closed-loop control calculation module, configured to perform closed-loop control calculations based on the filtered Hall sensor signal and the filtered power supply signal to obtain a non-linear PWM duty cycle control quantity and a phase distribution parameter;

[0009] A phase difference control module, configured to perform phase difference control based on the non-linear PWM duty cycle control quantity and the phase distribution parameter to obtain a multi-fan collaborative control strategy.

[0010] Optionally, in the first implementation manner of the present invention, the dual analysis module further includes:

[0011] An acquisition unit, configured to acquire a rotational speed feedback signal of a double ball bearing fan within a preset rotational speed range, perform spectrum analysis on the rotational speed feedback signal to obtain an interference frequency distribution map; perform peak detection and energy statistics on the interference frequency distribution map to obtain a main interference frequency and an interference energy proportion coefficient within each frequency band, and perform mutation point detection and periodic analysis on the rotational speed feedback signal in the time domain to obtain interference waveform characteristics and a time domain distribution pattern;

[0012] A parameter fitting unit, configured to construct a four-dimensional electromagnetic interference correlation function based on the interference frequency distribution map, the main interference frequency, the interference energy proportion coefficient, the interference waveform characteristics, and the time domain distribution pattern; perform parameter fitting on the four-dimensional electromagnetic interference correlation function under different temperature conditions to obtain a multi-dimensional electromagnetic interference characteristic model including rotational speed-interference mapping, temperature-frequency offset, voltage-common mode interference, and load-differential mode interference relationships;

[0013] A parameter configuration unit, configured to configure the filter network parameters of an initial three-stage cascaded dynamic filter network based on the multi-dimensional electromagnetic interference characteristic model to obtain a target three-stage cascaded dynamic filter network.

[0014] Optionally, in the second implementation manner of the present invention, the parameter configuration unit is specifically configured to:

[0015] Configure the parameters of a π-type LC power filter circuit according to the voltage-common mode interference relationship in the multi-dimensional electromagnetic interference characteristic model to obtain a first-stage filter circuit, where the first-stage filter circuit includes a common mode choke coil and a low ESR electrolytic capacitor;

[0016] Calculate the center frequency and bandwidth of a band-pass filter based on the rotational speed-interference mapping and temperature-frequency offset relationships in the multi-dimensional electromagnetic interference characteristic model to obtain a second-stage signal conditioning filter circuit;

[0017] Optimize and calculate the coefficient matrix of a digital filter based on the load-differential mode interference relationship in the multi-dimensional electromagnetic interference characteristic model to obtain a third-stage digital filter circuit;

[0018] Cascade connect the first - stage filtering circuit, the second - stage signal conditioning and filtering circuit, and the third - stage digital filtering circuit through an opto - isolator to obtain an initial three - stage cascaded dynamic filtering network;

[0019] Test the transfer function of the initial three - stage cascaded dynamic filtering network under simulated electromagnetic interference conditions to obtain the actual filtering characteristic curve and error parameters;

[0020] Adjust the parameters of the initial three - stage cascaded dynamic filtering network according to the actual filtering characteristic curve and the error parameters to obtain a target three - stage cascaded dynamic filtering network.

[0021] Optionally, in the third implementation mode of the present invention, the signal purification module is specifically used for:

[0022] Collect the first fan operation parameters from the Hall sensor, power supply monitoring port, temperature sensor, and current sampling circuit of the double - ball bearing fan;

[0023] Input the power supply voltage waveform in the first fan operation parameters into the first - stage filtering circuit of the target three - stage cascaded dynamic filtering network for common - mode interference suppression to obtain an initial power supply signal;

[0024] Input the rotational speed pulse signal in the first fan operation parameters into the second - stage signal conditioning and filtering circuit of the target three - stage cascaded dynamic filtering network for band - pass filtering and automatic gain control to obtain an initial Hall sensor signal;

[0025] Perform temperature compensation calculation on the gain parameter of the second - stage signal conditioning and filtering circuit based on the ambient temperature value in the first fan operation parameters to obtain a corrected signal gain coefficient;

[0026] Input the initial power supply signal and the initial Hall sensor signal into the third - stage digital filtering circuit of the target three - stage cascaded dynamic filtering network for adaptive IIR filtering processing according to the corrected signal gain coefficient, and dynamically adjust the filter order according to the motor current value in the first fan operation parameters to obtain a filtered Hall sensor signal and a filtered power supply signal.

[0027] Optionally, in the fourth implementation mode of the present invention, the closed - loop control calculation module is specifically used for:

[0028] Perform edge detection and period calculation on the filtered Hall sensor signal to obtain the actual rotational speed value and rotational speed stability factor, and compare and calculate the actual rotational speed value with a preset rotational speed target value to obtain a rotational speed error value and a rotational speed change rate;

[0029] Based on the rotational speed error value, the rotational speed change rate, and the rotational speed stability factor, calculate the PID control parameters to obtain a linear control quantity;

[0030] Input the linear control quantity into a piecewise linear mapping function for non-linear transformation processing to obtain a non-linear PWM duty cycle control quantity. The piecewise linear mapping function adopts different slope values in different rotational speed intervals;

[0031] Based on the number and physical layout information of the double ball fans, perform phase optimization calculation to obtain phase distribution parameters.

[0032] Optionally, in the fifth implementation manner of the present invention, the phase difference control module is specifically used for:

[0033] Send the non-linear PWM duty cycle control quantity and the phase distribution parameters to each slave controller in the distributed control network through the RS-485 bus to obtain a control frame data packet;

[0034] Perform fan number self-adaptive calculation on the phase distribution parameters in the control frame data packet to obtain the phase difference between adjacent fans, and perform offset processing on the PWM signal generation timing of each slave controller based on the phase difference to obtain a multi-channel PWM drive signal;

[0035] Perform real-time monitoring and analysis on the current waveforms of the multi-channel PWM drive signals to obtain the total current ripple characteristics and the electromagnetic interference degree data between each fan, and perform matching selection on the working modes of each fan based on the total current ripple characteristics and the electromagnetic interference degree data to obtain a working mode allocation scheme;

[0036] Combine the working mode allocation scheme with the non-linear PWM duty cycle control quantity, and dynamically adjust the actual rotational speed allocation of each fan according to the load coefficient and heat dissipation requirements of each fan to obtain a multi-fan cooperative control strategy.

[0037] Optionally, in the sixth implementation manner of the present invention, the intelligent heat dissipation fan system with electromagnetic interference resistance based on a dynamic filter circuit design further includes:

[0038] An execution module, configured to execute the multi-fan cooperative control strategy and collect second fan operation parameters from the distributed fan control network. The second fan operation parameters include electrical parameters, mechanical parameters, and control parameters;

[0039] A multi-dimensional determination module is used to perform stability analysis and anomaly determination on the electrical parameters in the operating parameters of the second fan to obtain an electrical parameter health index. The electrical parameters include power supply voltage stability, motor current waveform, power factor, and insulation resistance; perform noise spectrum and vibration characteristic analysis on the mechanical parameters in the operating parameters of the second fan to obtain a mechanical parameter health index. The mechanical parameters include bearing noise, vibration amplitude, starting characteristics, and rotational speed stability; perform response characteristic and anti-interference ability analysis on the control parameters in the operating parameters of the second fan to obtain a control parameter health index. The control parameters include control accuracy, response time, anti-interference ability, and control margin;

[0040] A response output module is used to perform weighted fusion on the electrical parameter health index, the mechanical parameter health index, and the control parameter health index to obtain a multi-dimensional parameter health index, and perform hierarchical response processing based on the comparison result between the multi-dimensional parameter health index and a preset health threshold, and output a corresponding filter parameter recalibration instruction.

[0041] In summary, the technical solution provided by the present invention realizes high-precision control and anti-interference ability within a wide rotational speed range through a multi-dimensional electromagnetic interference characteristic model and a three-stage cascaded dynamic filtering network. The phase difference control strategy is adopted to solve the problem of mutual influence of electromagnetic interference when multiple fans are connected in parallel, effectively reducing the total current ripple and improving the system stability. Based on a multi-dimensional health assessment mechanism of electrical parameters, mechanical parameters, and control parameters, combined with hierarchical response processing, the self-diagnosis and self-repair functions of the fan system are realized, and the service life of the fan is extended. Through non-linear PWM duty cycle control and intelligent matching of working modes, the system can accurately adjust the air volume and air pressure according to the heat dissipation requirements while keeping the noise within a reasonable range. The system design takes into account dual-voltage compatibility, wide temperature range adaptability, and aging compensation functions, and can maintain stable control performance under complex electromagnetic environments and various working conditions, improving the energy utilization efficiency and reliability of the overall system. Description of the Drawings

[0042] Figure 1 It is a schematic diagram of an anti-electromagnetic interference intelligent heat dissipation fan system designed based on a dynamic filtering circuit in an embodiment of the present invention.

[0043] The realization, functional characteristics, and advantages of the purpose of the present invention will be further described in conjunction with the embodiments with reference to the drawings. Detailed Embodiments

[0044] In order to make the purpose, technical solution, and advantages of the present invention clearer, the present invention will be further described in detail below in conjunction with the drawings and embodiments. It should be understood that the specific embodiments described here are only used to explain the present invention and are not used to limit the present invention.

[0045] Reference Figure 1 , this embodiment provides an anti-electromagnetic interference intelligent cooling fan system based on the design of a dynamic filter circuit, including:

[0046] A dual-analysis module 1, configured to perform dual-time-domain and frequency-domain analysis on the rotational speed feedback signal of the double-ball bearing fan, obtain a multi-dimensional electromagnetic interference characteristic model, and configure the filter network parameters of the initial three-stage cascaded dynamic filter network to obtain a target three-stage cascaded dynamic filter network;

[0047] Among them, for an intelligent heat dissipation fan system with anti-electromagnetic interference based on a dynamic filtering circuit design and equipped with the functions of the acquisition unit, parameter fitting unit, and parameter configuration unit in the dual analysis module, a closed-loop system for signal acquisition, analysis, modeling, and optimal control needs to be constructed. Among them, the implementation of the acquisition unit depends on the signal acquisition circuit and the calculation module. During the operation of the fan, the rotational speed feedback signal of the double ball bearing fan within the preset rotational speed range is collected in real time, and it is ensured that the accuracy of signal acquisition is high enough to avoid inaccurate extraction of interference characteristics due to sampling errors. After being amplified at the front end and converted from analog to digital, the signal enters the signal processing unit, and the fast Fourier transform or short-time Fourier transform is used to perform spectral analysis on the rotational speed feedback signal to obtain the interference frequency distribution map. Peak detection is performed on the interference spectrum to extract the main interference frequencies within a specific frequency band, and the interference energy proportionality coefficient of each frequency band is calculated by combining the signal energy statistical method, so as to quantitatively determine the intensity of interference in different frequency bands. In order to further analyze the characteristics of interference in the time domain, the mutation point detection is performed on the rotational speed feedback signal to capture transient interference events, and periodic analysis is performed to extract the time domain distribution pattern of interference, so as to obtain the interference waveform characteristics. Through the parameter fitting unit, a four-dimensional electromagnetic interference correlation function is constructed and modeled for the interference frequency distribution map, main interference frequencies, interference energy proportionality coefficient, interference waveform characteristics, and time domain distribution pattern, and a four-dimensional electromagnetic interference correlation function is constructed. Using non-linear regression or neural network fitting methods, multivariate fitting is performed on these data to establish an accurate mathematical model. In order to improve the adaptability of the model under different environmental conditions, it is extended with working condition factors such as temperature, load, and voltage, and fitting under different temperature conditions is realized, so that the model reflects the influence of different environmental variables on the electromagnetic interference characteristics of the fan. For example, the rotational speed-interference mapping relationship is introduced to reveal the changes in electromagnetic interference characteristics at different rotational speeds, and at the same time, the temperature-frequency offset relationship is established to reflect the influence of the fan operating temperature on the interference spectrum. The voltage-common mode interference relationship is considered to quantify the influence of the supply voltage change on the common mode noise, and the load-differential mode interference relationship is established to describe the contribution of the fan load fluctuation to the differential mode interference. Through this step, a multi-dimensional electromagnetic interference characteristic model that comprehensively describes the electromagnetic interference characteristics is constructed. Using the parameter configuration unit, based on the established multi-dimensional electromagnetic interference characteristic model, the filtering parameters of the initial three-stage cascaded dynamic filtering network are dynamically adjusted to ensure that the filtering network optimally suppresses the electromagnetic interference under different working conditions. The design of the three-stage cascaded dynamic filtering network includes three levels: active filtering, adaptive notch filtering, and low-pass filtering. Among them, the active filtering part is used to suppress high-frequency interference, and the adaptive notch filtering is used to dynamically adjust the notch frequency to match the main interference frequency and effectively reduce the amplitude of the interference signal, while the low-pass filtering part is used to smooth the signal and remove high-frequency noise.Based on the multi-dimensional electromagnetic interference feature model, optimize the filter parameters under different rotational speeds, temperatures, voltages, and load conditions to achieve a more targeted dynamic anti-interference ability, and obtain the target three-stage cascaded dynamic filtering network.

[0048] Configure the parameters of the π-type LC power filter circuit according to the voltage-common mode interference relationship to construct the first-stage filter circuit. This circuit uses a common mode choke to suppress high-frequency common mode noise and combines low-ESR (equivalent series resistance) electrolytic capacitors to filter out low-frequency fluctuations to ensure the power supply stability of the fan. The inductance value of the common mode choke is determined by considering the power supply voltage and the characteristics of common mode interference, while the ESR parameter of the electrolytic capacitor needs to be as low as possible to reduce the impact on the power supply signal and improve the filtering effect. In this way, the first-stage filter circuit effectively reduces the common mode interference at the power supply end, thereby reducing the impact of EMI on the entire system. Calculate the center frequency and bandwidth of the band-pass filter based on the speed-interference mapping and temperature-frequency offset relationship in the multi-dimensional electromagnetic interference characteristic model to construct the second-stage signal conditioning and filtering circuit. Since the speed of the fan directly affects the distribution of interference frequencies, and temperature changes will cause changes in material properties, which in turn affect the frequency offset of interference signals, the center frequency of the band-pass filter cannot be a fixed value but needs to be dynamically adjusted according to the operating state of the fan. By analyzing the speed-interference mapping relationship, determine the distribution range of the main interference frequencies at different fan speeds. At the same time, combine the temperature-frequency offset relationship to establish a temperature compensation mechanism to keep the band-pass filter in the best frequency matching when the temperature changes. The calculation of the center frequency uses an adaptive algorithm to adjust the filter parameters according to the real-time monitored speed and temperature, so that it always provides the best suppression effect for the main interference frequency band. At the same time, the selection of the bandwidth takes into account both the filtering effect and signal fidelity to ensure that the filtering does not cause excessive attenuation of the normal fan control signal, thereby affecting the stability of the system. Optimize the coefficient matrix of the digital filter based on the load-differential mode interference relationship to obtain the third-stage digital filter circuit. Since the change in the fan load will cause the dynamic characteristics of the differential mode interference to change, the digital filter needs to have the ability of adaptive adjustment to provide efficient filtering effect under different load conditions. Use FIR (finite impulse response) or IIR (infinite impulse response) filters and dynamically adjust the filter coefficients through online calculation to perform optimal filtering for the current interference characteristics. Construct an optimization algorithm based on the load-differential mode interference relationship. This algorithm calculates the optimal filter coefficient matrix according to the real-time monitored load change and performs real-time adjustment through a digital signal processor or FPGA, so as to ensure that the filter is always in the best working state. To avoid introducing additional delay to the system by the digital filter, optimize the calculation efficiency during filter design and select an appropriate filtering order to balance the filtering performance and computational complexity. Connect the first-stage filter circuit, the second-stage signal conditioning and filtering circuit, and the third-stage digital filter circuit in cascade through an opto-isolator to construct an initial three-stage cascade dynamic filtering network. The role of the opto-isolator is to prevent interference coupling between different filtering stages and ensure that the signal is not affected by external noise during the filtering process.The transfer function of the initial three-stage cascaded dynamic filtering network is tested to evaluate its actual filtering performance and ensure that it meets the anti-interference requirements of the system. The testing process includes measuring the input and output signals of the filtering network under simulated electromagnetic interference conditions and calculating the actual filtering characteristic curve through Fourier transform. By comparing the signal spectra before and after filtering, the attenuation characteristics of each filtering stage are analyzed, and the error parameters are calculated to evaluate whether the actual performance of the filter meets the design expectations. If there is a large deviation between the actual filtering characteristic curve and the ideal filtering characteristic curve calculated theoretically, the parameters of the initial three-stage cascaded dynamic filtering network are adjusted to correct the operating state of the filter. For example, the inductance value or capacitance value of the π-type LC filter is adjusted to optimize the suppression effect of common-mode interference; or the center frequency and bandwidth of the band-pass filter are adjusted to improve the matching degree for the target interference frequency band; or the coefficient matrix of the digital filter is re-optimized to improve the adaptability of the filter to different load conditions. Through iterative optimization, the target three-stage cascaded dynamic filtering network is obtained.

[0049] The signal purification module 2 is configured to collect the first fan operation parameters of the double ball bearing fan and input them into the target three-stage cascaded dynamic filtering network for signal purification, so as to obtain the filtered Hall sensor signal and the filtered power supply signal;

[0050] Specifically, the first fan operation parameters are collected from the Hall sensor, power supply monitoring port, temperature sensor, and current sampling circuit of the double-ball bearing fan. The Hall sensor provides a pulse signal of the fan speed. The power supply monitoring port is used to measure the working voltage waveform of the fan. The temperature sensor is used to detect the temperature of the environment where the fan is located. The current sampling circuit is used to obtain the operating current of the fan. The acquisition of these data is performed by a high-precision analog-to-digital conversion (ADC) module for signal sampling to ensure the accuracy of the data, and the input signal is preprocessed by a low-noise operational amplifier to enhance the signal quality and reduce the impact of initial interference. The power supply voltage waveform is input into the first-stage filtering circuit of the target three-stage cascaded dynamic filtering network. This circuit adopts a π-type LC filtering structure, where the common-mode choke is used to suppress high-frequency common-mode noise, and the low-ESR electrolytic capacitor is used to filter out low-frequency fluctuation signals to ensure the power supply quality of the fan. During this process, the filtering network dynamically adjusts the inductance value of the common-mode choke and the capacitance of the electrolytic capacitor by analyzing the voltage-common-mode interference relationship in the multi-dimensional electromagnetic interference characteristic model to adapt to the power supply interference characteristics under different operating states. After being processed by this filtering stage, an initial power supply signal is obtained. At the same time, the rotation speed pulse signal in the first fan operation parameters is input into the second-stage signal conditioning and filtering circuit of the target three-stage cascaded dynamic filtering network. This filtering stage adopts band-pass filtering and automatic gain control to ensure the stability of the Hall sensor signal. During this process, the center frequency of the band-pass filter is dynamically adjusted according to the rotation speed-interference mapping relationship to match the signal frequency corresponding to the current rotation speed of the fan, thereby effectively filtering out high-frequency noise and low-frequency drift signals. At the same time, to cope with the fluctuation of the amplitude of the Hall sensor signal, the automatic gain control module automatically adjusts the gain parameter according to the amplitude of the input signal, so that the Hall signal always remains within an appropriate amplitude range to ensure the consistency and reliability of the signal amplitude during subsequent processing. After being processed by this filtering stage, an initial Hall sensor signal is obtained. Temperature compensation calculation is performed on the gain parameter of the second-stage signal conditioning and filtering circuit based on the ambient temperature value in the first fan operation parameters to improve the adaptability of the filtering network under different temperature conditions. Since the change in temperature will affect the characteristics of electronic components, especially the response characteristics of operational amplifiers, filtering capacitors, and Hall sensors, the ambient temperature value in the first fan operation parameters is collected by the temperature sensor, and the corrected signal gain coefficient is calculated according to the temperature-frequency offset relationship to compensate for the gain deviation caused by temperature changes. The initial power supply signal and the initial Hall sensor signal are input into the third-stage digital filtering circuit of the target three-stage cascaded dynamic filtering network, and the residual electromagnetic interference is suppressed through adaptive IIR filtering. The IIR filter provides high filtering performance with low computational complexity and adapts to the interference characteristics under different operating conditions by updating the filter coefficients in real time.During this process, the order of the filter is dynamically adjusted according to the motor current value in the first fan operating parameter, because the change in the motor current will affect the load characteristics of the fan, thereby changing the amplitude and spectral distribution of the differential-mode interference. By real-time monitoring the operating current of the fan and calculating the optimal filter order based on the load-differential-mode interference relationship, the computational burden is reduced while ensuring the filtering performance, thereby improving the response speed and real-time performance of the system. After the processing of the third-stage digital filtering circuit, the filtered Hall sensor signal and the filtered power supply signal are obtained.

[0051] The closed-loop control calculation module 3 is used to perform closed-loop control calculations based on the filtered Hall sensor signal and the filtered power supply signal to obtain the non-linear PWM duty cycle control quantity and the phase distribution parameter;

[0052] It should be noted that edge detection is performed on the filtered Hall sensor signal to identify the rising or falling edge of the pulse signal, thereby calculating the pulse period and obtaining the actual fan speed value. At the same time, by analyzing the fluctuations within multiple consecutive periods, the rotational speed stability factor is calculated. This factor is used to measure the stability of the fan operation and serves as the basis for optimizing the control algorithm. The actual speed value is compared with the preset speed target value to obtain the speed error value. Meanwhile, the change rate of the fan speed within adjacent time periods is calculated to predict the changing trend of the fan operation state, and feedforward adjustment is introduced into the control algorithm to improve the response speed and accuracy of the control. Based on the speed error value, speed change rate, and speed stability factor, the PID control parameters are calculated to generate the basic linear control quantity. The proportional control part quickly adjusts the output according to the magnitude of the speed error. The integral control part eliminates long-term deviations by accumulating errors, while the derivative control part adjusts the output in advance by detecting the rate of speed change to reduce sudden changes in the fan speed. To enhance the adaptive ability of the system, the PID control parameters are dynamically adjusted in combination with the fan operation state, enabling the fan to achieve the best control effect under different load and environmental conditions, ensuring a smooth transition of the speed, and reducing the vibration and noise of the motor. The linear control quantity is input into a piecewise linear mapping function for nonlinear transformation processing to convert the linear control quantity into the final PWM signal. At low speeds, the PWM signal has a greater impact on the fan speed. Therefore, the mapping function adopts a smaller change amplitude to avoid speed fluctuations caused by over-control. At high speeds, the impact of the PWM signal is smaller, so the change amplitude is increased to improve the adjustment sensitivity to ensure that the fan quickly responds to changes in the external load. In the application scenario of multi-fan collaborative control, phase optimization calculation is performed based on the number and physical layout information of the fans to reduce the flow field interference and common-mode noise between the fans. Since multiple fans operating in the same system will cause chaotic airflows due to speed differences, reducing the overall heat dissipation efficiency, through phase optimization calculation, the PWM signals of different fans are phase-adjusted so that their operating rhythms are staggered from each other, reducing fluid disturbances and resonance phenomena. This optimization calculation combines the spatial layout, speed characteristics, and load conditions of the fans to ensure that the mutual interference between the fans is minimized, while maintaining a stable heat dissipation airflow, thereby improving the thermal management efficiency of the entire system and reducing the impact of electromagnetic interference on the control signal.

[0053] The phase difference control module 4 is used to perform phase difference control based on the non-linear PWM duty cycle control quantity and the phase distribution parameters to obtain the multi-fan collaborative control strategy.

[0054] Specifically, the non-linear PWM duty cycle control quantity and the phase distribution parameters are sent to each slave controller in the distributed control network via the RS-485 bus, so as to establish synchronous control among multiple fans. The RS-485 communication protocol has strong anti-interference ability and long-distance data transmission ability, which can ensure the reliable transmission of fan control signals in a complex electromagnetic environment. The master controller generates a control frame data packet containing the PWM control quantity and the phase distribution parameters based on the current operating state, load condition and speed requirement of the fan, and broadcasts it to all slave controllers via the RS-485 bus. After receiving the control frame, each slave controller analyzes the phase distribution parameters therein and adjusts them in combination with its own operating state to ensure the synchronous operation and phase optimization of multiple fans. The phase difference control module adaptively calculates the number of fans for the phase distribution parameters in the control frame data packet to determine the phase difference between adjacent fans. This calculation process involves factors such as the physical arrangement of the fans, the operating speed and the air flow direction, and it is necessary to allocate a reasonable phase shift among different fans to reduce the fluid interference and common-mode noise between the fans. Specifically, the phase shift angle is calculated according to the number and operating mode of the fans, so that the working cycles of adjacent fans are staggered as much as possible, avoiding the air flow disorder and the accumulation of electromagnetic interference caused by synchronous operation. After calculating the phase difference, the timing of the PWM signal generation of each slave controller is offset based on this phase difference to obtain multiple PWM drive signals, so that the PWM waveforms between different fans are adjusted according to the optimal phase relationship, realizing the efficient collaborative operation between the fans. The phase difference control module monitors and analyzes the current waveforms of multiple PWM drive signals in real time to obtain the total current ripple characteristics and the data of the electromagnetic interference degree between each fan. The analysis of the current waveform is to evaluate the overall power supply stability of the fan system and the influence of the common-mode noise between the fans. By performing high-frequency sampling on the current ripple of the PWM signal and combining digital signal processing technology, the main electromagnetic interference sources in the fan power supply system are extracted, and the current harmonic components between different fans are calculated. Based on these data, it is identified which fans have a large electromagnetic coupling, and corresponding anti-interference measures are taken, such as adjusting the duty cycle of the PWM signal, modifying the phase shift of the fan, or optimizing the filter parameters, to reduce the overall electromagnetic interference level. At the same time, based on the total current ripple characteristics and the electromagnetic interference degree data, the working modes of each fan are matched and selected, so as to formulate an optimal working mode allocation scheme. In this process, according to the load characteristics, temperature environment and current fluctuation of the fan, the most suitable operating mode is assigned to each fan, such as high-speed operation, low-speed operation, intermittent operation or energy-saving mode, to balance the heat dissipation capacity and power consumption requirements of the fan and minimize the electromagnetic interference. The working mode allocation scheme is combined with the non-linear PWM duty cycle control quantity, and the actual speed distribution of each fan is dynamically adjusted according to the load coefficient and heat dissipation requirement of each fan to optimize the efficiency and stability of the entire heat dissipation system.Combined with the load factor of each fan, the duty cycle of its PWM signal is adjusted in real time so that the rotation speed of each fan matches the current heat dissipation requirements while maintaining the power balance of the entire system. To reduce the electromagnetic interference between fans, the phase shift of each fan is dynamically adjusted according to the current ripple and interference data, so that they can avoid interfering with each other as much as possible during operation, thereby improving the reliability of the entire system.

[0055] The execution module executes the multi-fan collaborative control strategy, continuously collects the operating parameters of the second fan from the distributed fan control network, and ensures that the control instructions for the fan can be accurately executed. The operating parameters of the second fan include electrical parameters, mechanical parameters, and control parameters. Among them, the electrical parameters mainly involve the power supply status and current signal of the fan, the mechanical parameters involve the motion characteristics of the fan, and the control parameters are used to evaluate the response performance and anti-interference ability of the fan. To ensure the accuracy of data collection, the execution module extracts stable fan operating data based on the real-time monitored data stream, combined with the filtered voltage, current, and rotational speed signals, and transmits this data to the multi-dimensional determination module for analysis. After completing the data collection, the multi-dimensional determination module evaluates the health status of the operating parameters of the second fan to judge the current working condition of the fan and predict the fault trend. This module conducts stability analysis and anomaly determination on the electrical parameters to calculate the electrical parameter health index. The stability of the power supply voltage is judged by long-term fluctuation analysis to determine whether the power supply system is stable. The analysis of the motor current waveform identifies whether there are abnormal current surges or non-linear interferences. The monitoring of the power factor is used to judge whether the motor is operating efficiently, and the measurement of the insulation resistance is used to detect the safety of the motor winding and circuit. If the changes in these parameters exceed the set threshold, it indicates that there are problems such as unstable power supply, winding aging, or increased electromagnetic interference in the fan, which will affect the overall performance of the fan. The multi-dimensional determination module conducts noise spectrum and vibration characteristic analysis on the mechanical parameters to calculate the mechanical parameter health index. During the operation of the fan, the analysis of the bearing noise reflects the wear condition of the mechanical structure. The change in the vibration amplitude reveals the imbalance problem of the fan blades or bearings. The monitoring of the starting characteristics identifies abnormal phenomena during the acceleration process of the motor, and the analysis of the rotational speed stability judges whether the fan is affected by external interference or internal friction. By performing frequency-domain and time-domain analysis on these data, it is identified whether there are mechanical faults in the fan, such as bearing aging, blade deformation, or insufficient lubrication, and the operating life and maintenance requirements of the fan are evaluated. This module conducts response characteristic and anti-interference ability analysis on the control parameters to calculate the control parameter health index. The measurement of the control accuracy reflects the tracking ability of the fan to the target rotational speed. The analysis of the response time is used to judge the dynamic adjustment speed of the fan when the load changes, and the evaluation of the anti-interference ability judges the reliability of the control system by analyzing the stability of the fan in the electromagnetic environment. The calculation of the control margin can determine whether the current adjustment ability of the fan is sufficient to ensure that the system can still maintain normal operation under extreme conditions. Through these data analyses, it is evaluated whether the fan control system needs to optimize the control parameters to improve the overall system stability and response speed. When the multi-dimensional determination module completes the health assessment, the response output module performs weighted fusion on the electrical parameter health index, mechanical parameter health index, and control parameter health index to calculate the comprehensive multi-dimensional parameter health index.During the calculation process, the module assigns different weights according to the influence degree of different parameters on the operating state of the fan, so that the final health index can accurately reflect the overall operating condition of the fan. The calculated multi-dimensional parameter health index is compared with the preset health threshold to determine whether the fan is in a normal operating state. If the health index is lower than the health threshold, it indicates that the fan has a potential fault or a risk of performance degradation, and corresponding adjustment measures need to be taken. After determining the health state of the fan, the response output module performs hierarchical response processing to formulate corresponding optimization strategies. The response output module outputs a filter parameter recalibration instruction to adaptively adjust the dynamic filter circuit of the fan, thereby optimizing the anti-interference ability of the fan. This instruction includes adjusting the capacitance value of the LC filter network to optimize the power supply stability, recalculating the center frequency of the band-pass filter to match the current interference spectrum, or adjusting the coefficient matrix of the digital filter to enhance the signal processing ability.

[0056] In one example, the dual analysis module 1 further includes:

[0057] An acquisition unit, configured to acquire the rotational speed feedback signal of the double ball bearing fan within a preset rotational speed range, perform spectral analysis on the rotational speed feedback signal to obtain an interference frequency distribution map; perform peak detection and energy statistics on the interference frequency distribution map to obtain the main interference frequencies and interference energy proportionality coefficients within each frequency band, and perform mutation point detection and periodic analysis on the rotational speed feedback signal in the time domain to obtain interference waveform characteristics and time domain distribution patterns;

[0058] A parameter fitting unit, configured to construct a four-dimensional electromagnetic interference correlation function based on the interference frequency distribution map, the main interference frequencies, the interference energy proportionality coefficients, the interference waveform characteristics, and the time domain distribution patterns; perform parameter fitting on the four-dimensional electromagnetic interference correlation function under different temperature conditions to obtain a multi-dimensional electromagnetic interference characteristic model including rotational speed-interference mapping, temperature-frequency offset, voltage-common mode interference, and load-differential mode interference relationships;

[0059] A parameter configuration unit, configured to configure the filter network parameters of the initial three-stage cascaded dynamic filter network based on the multi-dimensional electromagnetic interference characteristic model to obtain a target three-stage cascaded dynamic filter network.

[0060] In this example, the acquisition unit obtains the rotational speed feedback signal from the double ball bearing fan, and this signal comes from a Hall sensor, whose pulse frequency is proportional to the actual rotational speed of the fan. Assume that the output signal of the Hall sensor is , and its expression form is:

[0061]

[0062] where represents the amplitude of the Hall signal, is the rotation frequency of the fan, is the initial phase of the signal, represents background noise. When analyzing this signal, fast Fourier transform is used for spectrum analysis to transform it into the frequency domain and obtain the interference frequency distribution map. For a discrete-time signal, the calculation formula of its fast Fourier transform is:

[0063]

[0064] where, represents the spectrum after transformation, is the sampling value of the discrete-time signal, is the window size of the fast Fourier transform, is the sampling period. Peak detection is performed on the interference spectrum, that is, to find the main interference frequencies with higher energy in the spectrum. Let the main interference frequency be , and the energy of these frequency components is statistically analyzed, and the interference energy proportion coefficient of each frequency band is calculated:

[0065]

[0066] where, represents the energy at the interference frequency , and the denominator part is the total energy of the entire spectrum. This ratio reflects the proportion of a specific interference frequency in the overall signal, thereby quantifying the influence degree of the main interference component. Analyze the signal in the time domain, including mutation point detection and periodicity analysis. The purpose of mutation point detection is to identify abnormal interference events during the operation of the fan, such as pulse signal distortion caused by external electromagnetic interference. It is achieved by calculating the first derivative of the signal and finding the mutation points:

[0067]

[0068] If shows abnormal peaks at certain time points, it indicates that electromagnetic interference mutation has occurred at that moment. At the same time, periodicity analysis is achieved through the autocorrelation function:

[0069]

[0070] where, the peak at corresponds to the periodic component of the fan speed, and non-periodic interference signals will not produce obvious correlation. By analyzing the shape of, identify the time-domain distribution pattern of the interference signal. The parameter fitting unit constructs a four-dimensional electromagnetic interference correlation function based on the obtained interference frequency distribution map, main interference frequency, interference energy proportion coefficient, interference waveform characteristics and time-domain distribution pattern, which is defined as:

[0071]

[0072] Among them, represents the rotational speed, represents the ambient temperature, is the power supply voltage, represents the fan load, is the parameter to be fitted, is the residual term. To make this function applicable to different working conditions, parameter fitting is carried out under different temperature conditions to establish a multi-dimensional electromagnetic interference characteristic model including rotational speed-interference mapping, temperature-frequency offset, voltage-common mode interference, and load-differential mode interference relationships. For example, when the temperature rises, the resistance of the motor coil increases, resulting in an offset of the electromagnetic interference frequency. This relationship is expressed as:

[0073]

[0074] Among them, is the temperature at which the interference frequency is is the reference temperature at which the interference frequency is is the frequency offset coefficient. Similarly, the change in voltage will affect the intensity of the common mode interference, and this relationship is fitted with a power function:

[0075]

[0076] Among them, is the amplitude of the common mode interference, is the fitting coefficient, represents the non-linear influence factor of the power supply noise. Through these parameter fittings, an electromagnetic interference characteristic model is established for subsequent filter optimization. After the parameter fitting is completed, the parameter configuration unit configures the filter parameters of the initial three-stage cascaded dynamic filter network based on this multi-dimensional electromagnetic interference characteristic model to optimize the anti-electromagnetic interference ability. The first-stage filter is mainly used for suppressing the common mode interference of the power supply signal, and its design is based on changes, and the filter transfer function is:

[0077]

[0078] Among them, and are the resistance and capacitance values of the filter circuit, which need to be optimized and adjusted according to amplitude. The second-stage filter is responsible for band-pass filtering to match the interference frequency related to the fan rotational speed, and its center frequency is calculated by , and the design of the band-pass filter satisfies:

[0079]

[0080] Among them, is the center frequency, is the filter quality factor. The third-stage filter uses adaptive IIR filtering to process the signal in real time, and its coefficients are optimized and adjusted according to the load variation. Through the processes of data acquisition, spectrum analysis, model fitting, and filter parameter optimization, it is ensured that the fan system operates stably in a complex electromagnetic environment, and the anti-interference ability is dynamically adjusted through adaptive filtering to improve the reliability and service life of the fan.

[0081] In one example, the parameter configuration unit is specifically used for:

[0082] Configuring the parameters of the π-type LC power filter circuit according to the voltage-common mode interference relationship in the multi-dimensional electromagnetic interference characteristic model to obtain the first-stage filter circuit, and the first-stage filter circuit includes a common mode choke and a low-ESR electrolytic capacitor;

[0083] Calculating the center frequency and bandwidth of the band-pass filter based on the speed-interference mapping and temperature-frequency offset relationship in the multi-dimensional electromagnetic interference characteristic model to obtain the second-stage signal conditioning and filtering circuit;

[0084] Optimizing and calculating the coefficient matrix of the digital filter based on the load-differential mode interference relationship in the multi-dimensional electromagnetic interference characteristic model to obtain the third-stage digital filter circuit;

[0085] Cascading and connecting the first-stage filter circuit, the second-stage signal conditioning and filtering circuit, and the third-stage digital filter circuit through an opto-isolator to obtain an initial three-stage cascaded dynamic filter network;

[0086] Testing the transfer function of the initial three-stage cascaded dynamic filter network under simulated electromagnetic interference conditions to obtain the actual filter characteristic curve and error parameters;

[0087] Adjusting the parameters of the initial three-stage cascaded dynamic filter network according to the actual filter characteristic curve and error parameters to obtain the target three-stage cascaded dynamic filter network.

[0088] In this example, when constructing the first-stage filter circuit, the parameters of the π-type LC power filter circuit are configured according to the voltage-common mode interference relationship in the multi-dimensional electromagnetic interference characteristic model. The common mode interference is mainly caused by the high-frequency noise of the power supply line, and its equivalent circuit is represented as a high-frequency voltage source superimposed on the DC supply voltage , the impedance of the interference source is represented as , and the equivalent model of the filter circuit is as follows:

[0089]

[0090] Among them, is the inductance value of the common-mode choke, is the equivalent series impedance of the power line, represents the Laplace transform variable. To optimize the filtering effect, select appropriate and low-ESR electrolytic capacitors to ensure that the common-mode interference is effectively attenuated within the target frequency band. The capacitance value is calculated based on the frequency distribution of the common-mode interference , and its cut-off frequency is determined by the following formula:

[0091]

[0092] According to the interference spectrum data, set lower than the main common-mode interference frequency, so as to ensure that the power supply noise is effectively filtered. Configure the second-stage signal conditioning and filtering circuit according to the speed-interference mapping and temperature-frequency offset relationship. This filtering circuit uses a band-pass filter to accurately match the main interference frequency under the operating state of the fan. The rotational speed of the fan and the electromagnetic interference frequency it generates are obtained through experimental measurement and are expressed as:

[0093]

[0094] Among them, and are fitting coefficients, which are affected by the structural characteristics of the fan. The temperature change will cause the change of the material resistance, thus affecting the frequency offset of the electromagnetic interference:

[0095]

[0096] Among them, is the temperature compensation coefficient. The center frequency of the band-pass filter is dynamically adjusted according to the current rotational speed and temperature to make it satisfy:

[0097]

[0098] The transfer function of the band-pass filter is expressed as:

[0099]

[0100] Among them, is the center frequency, is the quality factor of the filter, which determines the size of the bandwidth. To ensure the matching accuracy of the band-pass filter, for Perform real-time adjustment. Calculate the coefficient matrix of the digital filter based on the load-differential mode interference relationship to obtain the third-stage digital filtering circuit. The differential mode interference comes from the non-linear load of the fan motor, manifested as current waveform distortion, and its equivalent mathematical model is expressed as:

[0101]

[0102] Among them, is the differential mode interference current, is the DC current component, is the interference amplitude, is the interference frequency. The digital filter adopts the adaptive IIR filtering method, and the transfer function is expressed as:

[0103]

[0104] Among them, and are the filter coefficients, and are the numerator and denominator orders respectively. To optimize the filtering performance, calculate the optimal filtering order according to the current load, so that the interference signal is attenuated to the maximum extent. For example, when the load current is large, the interference frequency will fluctuate with the current, resulting in the need to increase the optimal order of the IIR filter to improve the filtering accuracy. After completing the design of the three-stage filter, cascade connection is carried out through an optocoupler isolator to prevent signal interference between different filtering stages. The role of the optocoupler isolator is to isolate the analog circuit part from the digital signal processing part to improve the anti-interference ability of the system and reduce the coupling of common-mode noise. To verify the performance of the filter, conduct a transfer function test in an analog electromagnetic interference environment, that is, measure the filtered output signal when a known interference signal is input, and calculate its frequency response characteristics. In the experiment, compare the signal power spectral density before and after filtering:

[0105]

[0106] Among them, and are the power spectral densities of the input and output signals respectively, and the actual attenuation ability of the filter is determined through this test. Based on the results of the transfer function test, calculate the error parameters of the filter and adjust the parameters of each filtering circuit to optimize the filtering performance. For example, if the test results show that the cut-off frequency of the common-mode filtering circuit is slightly higher than the target value, appropriately increase the inductance value of the common-mode choke to enhance the low-frequency interference suppression effect; if the center frequency of the band-pass filter deviates greatly, adjust the temperature compensation coefficient to optimize the filtering matching; if the performance of the digital filter is not stable enough, increase the filtering order to improve the dynamic response ability. After multiple iterations of optimization, finally obtain the target three-stage cascaded dynamic filtering network.

[0107] In one example, the signal purification module 2 is specifically configured to:

[0108] Collect first fan operation parameters from the Hall sensor, power supply monitoring port, temperature sensor, and current sampling circuit of the double ball bearing fan;

[0109] Input the power supply voltage waveform in the first fan operation parameters into the first-stage filtering circuit of the target three-stage cascaded dynamic filtering network for common-mode interference suppression to obtain an initial power supply signal;

[0110] Input the rotation speed pulse signal in the first fan operation parameters into the second-stage signal conditioning and filtering circuit of the target three-stage cascaded dynamic filtering network for band-pass filtering and automatic gain control to obtain an initial Hall sensor signal;

[0111] Perform temperature compensation calculation on the gain parameter of the second-stage signal conditioning and filtering circuit based on the ambient temperature value in the first fan operation parameters to obtain a corrected signal gain coefficient;

[0112] Input the initial power supply signal and the initial Hall sensor signal into the third-stage digital filtering circuit of the target three-stage cascaded dynamic filtering network for adaptive IIR filtering processing according to the corrected signal gain coefficient, and dynamically adjust the filter order according to the motor current value in the first fan operation parameters to obtain a filtered Hall sensor signal and a filtered power supply signal.

[0113] In this example, collect first fan operation parameters from the Hall sensor, power supply monitoring port, temperature sensor, and current sampling circuit of the double ball bearing fan. Among them, the Hall sensor provides a pulse signal of the fan rotation speed, and this signal is usually periodic and is defined as:

[0114]

[0115] Wherein, is the Hall signal amplitude, is the fan rotation speed frequency, is the signal phase, represents noise interference. At the same time, the power supply monitoring port collects the fan power supply voltage signal , this signal is affected by power grid fluctuations and high-frequency interference, while the temperature sensor measures the ambient temperature , and the current sampling circuit obtains the operating current of the fan motor , which is used to evaluate the motor load condition. After the data acquisition is completed, input the power supply voltage waveform in the first fan operation parameters into the first-stage filtering circuit of the target three-stage cascaded dynamic filtering network to perform common-mode interference suppression to obtain an initial power supply signal. The common-mode interference comes from the high-frequency noise of the power supply line, and its input signal model is expressed as:

[0116]

[0117] Among them, is the power supply signal affected by interference, representing the common-mode noise component. In order to filter out the common-mode interference, the first-stage filtering circuit adopts type LC filter, and its transfer function is expressed as:

[0118]

[0119] Among them, is the inductance value of the common-mode choke, is the equivalent impedance of the circuit. The common-mode choke is combined with the low-ESR electrolytic capacitor to form a common-mode filtering network, so that the high-frequency interference components are shunted in the capacitor, thereby reducing the impact of common-mode interference. The cut-off frequency needs to be less than the main common-mode interference frequency, satisfying:

[0120]

[0121] Input the rotation speed pulse signal in the first fan operation parameter into the second-stage signal conditioning and filtering circuit of the target three-stage cascaded dynamic filtering network. This filtering circuit adopts a band-pass filtering and automatic gain control mechanism to enhance the stability of the fan rotation speed feedback signal. The frequency of the fan rotation speed signal is affected by the rotation speed-interference mapping relationship and is expressed as:

[0122]

[0123] Among them, is the interference frequency, is the angular velocity of the fan, and are the fitting coefficients. The center frequency of the band-pass filter needs to be dynamically adjusted to satisfy:

[0124]

[0125] Among them, is the temperature drift coefficient, is the ambient temperature. The transfer function of the band-pass filter is:

[0126]

[0127] Among them, , is the quality factor of the filter. The automatic gain control mechanism is used to dynamically adjust the gain according to the amplitude of the input signal, satisfying:

[0128]

[0129] Among them, is the target signal amplitude, is the measured amplitude of the current signal, ensuring that the signal amplitude can remain within a stable range regardless of how the fan speed changes. Based on the ambient temperature value in the first fan operating parameter, temperature compensation calculation is performed on the gain parameter of the second-stage signal conditioning and filtering circuit to obtain the corrected signal gain coefficient. Since temperature changes will affect the gain characteristics of electronic components, the corrected gain parameter satisfies:

[0130]

[0131] Among them, is the temperature compensation factor. According to the corrected signal gain coefficient, the initial power supply signal and the initial Hall sensor signal are input into the third-stage digital filtering circuit of the target three-stage cascaded dynamic filtering network, and adaptive IIR filtering processing is adopted. At the same time, the filter order is dynamically adjusted according to the motor current value in the first fan operating parameter. The transfer function of the IIR filter is:

[0132]

[0133] Among them, and are the filter coefficients, and are the filter orders. Since the fan current waveform is affected by the load, when the motor load increases, the interference frequency will change accordingly. Therefore, the filter order needs to be dynamically adjusted to satisfy:

[0134]

[0135] Among them, is the adjusted filter order, is the reference current, is the load compensation factor. For example, under high-load conditions, the filter order is increased to enhance the filtering accuracy, ensuring that regardless of the operating state of the fan, interference can be suppressed to the greatest extent. The filtered Hall sensor signal and power supply signal are transmitted to the control system to ensure the stable operation of the fan in a complex electromagnetic environment and improve the overall heat dissipation efficiency and control accuracy.

[0136] In one example, the closed-loop control calculation module 3 is specifically used for:

[0137] Performing edge detection and period calculation on the filtered Hall sensor signal to obtain the actual rotation speed value and the rotation speed stability factor, and comparing and calculating the actual rotation speed value with the preset rotation speed target value to obtain the rotation speed error value and the rotation speed change rate;

[0138] PID control parameters are calculated based on the rotational speed error value, the rate of change of rotational speed, and the rotational speed stability factor to obtain a linear control quantity;

[0139] The linear control quantity is input into a piecewise linear mapping function for non-linear transformation processing to obtain a non-linear PWM duty cycle control quantity. The piecewise linear mapping function uses different slope values in different rotational speed intervals;

[0140] Phase optimization calculation is performed based on the number and physical layout information of the double ball fans to obtain phase distribution parameters.

[0141] In this example, edge detection and period calculation are performed on the filtered Hall sensor signal to obtain the actual rotational speed value of the fan. The signal output by the Hall sensor is a periodic pulse signal, and its mathematical expression is:

[0142]

[0143] where, is the signal amplitude, is the rotational frequency of the fan, is the phase offset, represents the sign function. To calculate the rotational speed of the fan, the rising edge or falling edge of the Hall signal is detected, and the time interval between pulses is measured ,then the actual rotational speed value is calculated by the formula:

[0144]

[0145] where, is the number of pulses output by the Hall sensor per revolution, is the time interval between two consecutive pulses detected. To evaluate the rotational speed stability of the fan, the rotational speed stability factor is calculated, and its definition is the ratio of the standard deviation of the rotational speed to the average rotational speed, that is:

[0146]

[0147] where, is the standard deviation of the rotational speed measured over multiple periods, is the average rotational speed. A smaller indicates that the fan runs stably, while a larger means that the load fluctuation or electromagnetic interference causes the rotational speed of the fan to change violently. The actual rotational speed value is compared with the preset rotational speed target value to calculate the rotational speed error value and the rate of change of rotational speed :

[0148]

[0149]

[0150] Among them, reflects the deviation between the current rotational speed and the target rotational speed of the fan, while reflects the dynamic change trend of the fan rotational speed. In order to enable the fan to quickly and stably adjust to the target rotational speed, based on , and perform PID control parameter calculation to obtain the linear control quantity . The standard formula for PID control is:

[0151]

[0152] Among them, is the proportional gain, is the integral gain, is the derivative gain. The proportional term is used to provide fast adjustment ability, the integral term is used to eliminate the steady-state error, and the derivative term predicts the rotational speed change trend to reduce overshoot and oscillation of the fan. The PID parameters are adaptively adjusted according to the fan operating state. For example, when is larger, appropriately increase to improve the dynamic response ability of the system, thereby suppressing the unstable oscillation of the fan. Perform a non-linear transformation on the linear control quantity output by PID control to optimize the control effect of the PWM signal. Input into the piecewise linear mapping function to obtain the non-linear PWM duty ratio control quantity . The purpose of non-linear mapping is to adopt different slope values in different rotational speed intervals, so that the fan has a smaller gain at low speeds and a larger gain at high speeds, thereby optimizing the dynamic response characteristics of the fan. The non-linear mapping function is expressed as:

[0153]

[0154] Among them, are the slopes in the low-speed, medium-speed, and high-speed intervals respectively, is the offset, is the interval division point. For example, in the low-speed interval, takes a smaller value to reduce the fluctuation of the PWM duty ratio and prevent low-speed jitter; while in the high-speed interval, takes a larger value to enhance the control sensitivity and improve the response speed. In the multi-fan collaborative control, perform phase optimization calculation based on the number and physical layout information of the double ball bearings fans to determine the phase distribution parameter Since the simultaneous operation of multiple fans can cause flow field interference and common-mode noise accumulation, the phase relationship of the fans is adjusted so that their operating rhythms are staggered to reduce the resonance effect and electromagnetic interference. The basic principle of phase optimization calculation is to stagger the phases of adjacent fans so that the conduction times of their PWM signals are as evenly distributed as possible throughout the cycle. The phase optimization relationship is expressed as:

[0155]

[0156] where is the total number of fans in the system, is the relative number of the fan.

[0157] In one example, the phase difference control module 4 is specifically used for:

[0158] Sending the non-linear PWM duty ratio control quantity and the phase distribution parameter to each slave controller in the distributed control network through the RS-485 bus to obtain a control frame data packet;

[0159] Performing fan number adaptive calculation on the phase distribution parameter in the control frame data packet to obtain the phase difference between adjacent fans, and performing offset processing on the PWM signal generation timing of each slave controller based on the phase difference to obtain a multiplexed PWM drive signal;

[0160] Performing real-time monitoring and analysis on the current waveforms of the multiplexed PWM drive signals to obtain the total current ripple characteristics and the electromagnetic interference degree data between each fan, and performing matching selection on the working modes of each fan based on the total current ripple characteristics and the electromagnetic interference degree data to obtain a working mode allocation scheme;

[0161] Combining the working mode allocation scheme with the non-linear PWM duty ratio control quantity, and dynamically adjusting the actual speed allocation of each fan according to the load coefficient and heat dissipation requirements of each fan to obtain a multi-fan cooperative control strategy.

[0162] In this example, the non-linear PWM duty ratio control quantity and the phase distribution parameter are sent to each slave controller in the distributed control network through the RS-485 bus to ensure that all fans can optimize phase synchronization during cooperative work and reduce common-mode noise interference. Calculate the non-linear PWM duty ratio control quantity of each fan , which is based on the linear control quantity output by the PID control Perform non-linear transformation through a piecewise linear mapping function, so that different gain coefficients are used in different speed intervals to optimize the response characteristics of the fan. The non-linear PWM duty ratio control quantity is expressed as:

[0163]

[0164] where is the PWM gain for different rotational speed ranges, is the offset, is the current fan rotational speed, and are the boundary values of the rotational speed range. Meanwhile, the phase distribution parameter of each fan is calculated according to the spatial arrangement of the fans to ensure that the PWM signals of adjacent fans are evenly distributed on the time axis, reducing air flow interference and common mode noise. The phase distribution calculation formula is:

[0165]

[0166] where, is the total number of fans, is the fan number. When the PWM control quantities and the phase parameters of all fans are calculated, these data are packed and sent to each slave controller through the RS-485 bus to form a control frame data packet. The RS-485 bus has strong anti-interference ability and allows multiple fan slave controllers to share the bus for data synchronization. The control frame data packet format is as follows:

[0167]

[0168] where, the fan ID is used to distinguish different slave controllers, is the temperature sensor data, is the current value, and CRC is used for data verification to prevent control failures caused by communication errors. After receiving the control frame data packet, each slave controller performs fan number adaptive calculation on the phase distribution parameter to calculate the phase difference between adjacent fans, and offsets the timing of the PWM signal to ensure that the PWM signals of the fans do not conduct simultaneously at the same time, so as to reduce the supply current ripple. The calculation of the phase difference between adjacent fans is as follows:

[0169]

[0170] Each slave controller adjusts the timing of the PWM signal based on this phase difference to make it evenly distributed within one cycle:

[0171]

[0172] where, is the PWM signal period, Number the current fan. This optimization ensures uniform phase distribution between fans, reduces current shock caused by simultaneous conduction, and thus reduces power supply noise. After completing the phase adjustment of the PWM signal, the current waveform of multiple PWM drive signals is monitored and analyzed in real time to calculate the total current ripple characteristics of the system. The degree of electromagnetic interference between the fans The total current ripple characteristics are calculated by Fourier transform:

[0173]

[0174] in, For the The current spectrum of each fan. The electromagnetic interference level is estimated by calculating the current correlation between the fans:

[0175]

[0176] in, Indicates fan and If If it is too large, it means that the electromagnetic coupling between fans is strong, and the phase or PWM duty cycle needs to be adjusted to reduce interference. Based on these analysis data, the fan working mode is matched to optimize system energy consumption and stability. The working mode allocation scheme is expressed as:

[0177]

[0178] in, Represents the fan's operating mode. The available modes include energy-saving mode (low-speed operation), high-efficiency mode (high-speed operation) or balanced mode (load balance of different fans). When the temperature is too high, it enters the balanced mode and reduces the number of fans working at the same time. When the current is too high, adjust the fan phase appropriately to reduce the current coupling interference. Combine the working mode allocation scheme with the nonlinear PWM duty cycle control amount and adjust the fan load factor according to the fan load factor. and heat dissipation requirements Dynamically adjust the actual speed distribution of each fan to optimize the overall cooling efficiency. The final fan speed is calculated as follows:

[0179]

[0180] in, Represents the cooling requirements of the fan area, is the fan load factor, which reflects the power consumption change of the fan under different load conditions. This optimization result is applied to all fans to obtain a multi-fan collaborative control strategy to ensure that the fans can operate in the optimal way under different load and temperature conditions.

[0181] In one example, the anti-electromagnetic interference intelligent cooling fan system based on a dynamic filtering circuit design further includes:

[0182] An execution module, configured to execute a multi-fan collaborative control strategy and collect second fan operation parameters from a distributed fan control network, where the second fan operation parameters include electrical parameters, mechanical parameters, and control parameters;

[0183] A multi-dimensional determination module, configured to perform stability analysis and anomaly determination on the electrical parameters in the second fan operation parameters to obtain an electrical parameter health index, where the electrical parameters include power supply voltage stability, motor current waveform, power factor, and insulation resistance; perform noise spectrum and vibration characteristic analysis on the mechanical parameters in the second fan operation parameters to obtain a mechanical parameter health index, where the mechanical parameters include bearing noise, vibration amplitude, starting characteristics, and rotational speed stability; perform response characteristic and anti-interference ability analysis on the control parameters in the second fan operation parameters to obtain a control parameter health index, where the control parameters include control accuracy, response time, anti-interference ability, and control margin;

[0184] A response output module, configured to perform weighted fusion on the electrical parameter health index, mechanical parameter health index, and control parameter health index to obtain a multi-dimensional parameter health index, and perform hierarchical response processing based on the comparison result between the multi-dimensional parameter health index and a preset health threshold, and output a corresponding filter parameter recalibration instruction.

[0185] In this example, the execution module executes a multi-fan collaborative control strategy to ensure that the fans operate normally according to the set non-linear PWM duty cycle control quantity and phase optimization parameters, and collects the second fan operation parameters in real time from the distributed fan control network. These parameters include electrical parameters, mechanical parameters, and control parameters. Among them, the electrical parameters reflect the power supply state and current characteristics of the fans, the mechanical parameters are used to evaluate the motion stability of the fans, and the control parameters are used to analyze the response ability and anti-interference performance of the fans. For the electrical parameters of the second fan, the power supply voltage stability is monitored in real time , which is defined as the ratio of the standard deviation to the average value of the power supply voltage within a short period of time:

[0186]

[0187] where represents the standard deviation of the power supply voltage, represents the average value of the power supply voltage. A higher means that there are large fluctuations or harmonic interferences in the power supply, which will affect the stability of the fans. Perform Fourier transform on the motor current waveform to extract the main harmonic components:

[0188]

[0189] If the energy on the non-fundamental frequency component is high, it indicates that the fan current is strongly harmonically interfered. Calculate the power factor to evaluate the energy conversion efficiency of the motor:

[0190]

[0191] wherein, is the active power, is the apparent power. When is much lower than 1, it indicates that the motor operating efficiency is low and there is excessive reactive power loss. The insulation resistance of the motor needs to be measured regularly to ensure the integrity of the winding insulation:

[0192]

[0193] wherein, is the applied test voltage, is the leakage current. If decreases too fast, it indicates that the insulation is aging or damp, resulting in enhanced electromagnetic interference. In addition to electrical parameters, analyze the mechanical parameters of the fan to evaluate the stability of its physical state. The bearing noise is obtained by a noise sensor and subjected to spectrum analysis:

[0194]

[0195] If there is a peak at a specific frequency, it indicates that the bearing is worn or lacks lubrication. The vibration amplitude is measured by an acceleration sensor and the vibration intensity is calculated through root mean square:

[0196]

[0197] Higher means that the fan is unbalanced or there is a bearing fault. The starting characteristics of the fan are calculated by measuring the ratio between the starting time and the time to reach the rated speed :

[0198]

[0199] Larger indicates that the motor starting process is blocked, caused by power supply problems or bearing jamming. The rotational speed stability of the fan is evaluated by calculating the rotational speed fluctuation amount:

[0200]

[0201] Higher means that the fan is affected by load fluctuations or electromagnetic interference, resulting in unstable operation. Based on the analysis of electrical and mechanical parameters, the control parameters of the fan are evaluated to calculate its health status. The control precision of the fan is calculated from the error between its actual rotational speed and the target rotational speed :

[0202]

[0203] Larger indicates that the control system cannot accurately track the target rotational speed. The response time reflects the time required for the fan to adjust from the initial rotational speed to the target rotational speed :

[0204]

[0205] Smaller indicates that the system has a stronger response ability. The anti-interference ability is calculated from the degree of fluctuation of the fan rotational speed during an electromagnetic interference event:

[0206]

[0207] Smaller indicates that the fan can resist external interference more stably. The control margin calculates the change range of the PWM duty cycle of the fan:

[0208]

[0209] where and are the maximum and minimum PWM duty cycles, is the rated duty cycle. Larger indicates that the control system still has a large adjustment space, while smaller indicates that the system is approaching saturation. When the health indices of electrical, mechanical, and control parameters are calculated, the response output module performs weighted fusion on these indices to obtain the multi-dimensional parameter health index of the fan :

[0210]

[0211] where is the weighting coefficient, are the health indices of electrical, mechanical, and control parameters respectively. Compare with the preset health threshold Compare. If , perform hierarchical response processing and adjust the filtering parameters of the fan. For example, when is slightly lower than the threshold, optimize the IIR filter coefficients to enhance the anti-interference ability; if decreases significantly, adjust the LC filter parameters to improve the power supply stability; if is far lower than the threshold, trigger a fan replacement recommendation. This system can monitor the operating status of the fan in real time and ensure the long-term stable operation of the fan through adaptive filtering adjustment and health assessment mechanisms.

[0212] The intelligent cooling fan system with anti-electromagnetic interference based on a dynamic filtering circuit design further includes:

[0213] An adaptive control module for real-time monitoring of the resonance characteristics of the fan system under the multi-fan cooperative control strategy to obtain system resonance frequency point data and harmonic amplitude value data; comparing and analyzing the system resonance frequency point data with the 1 / 6 threshold of the PWM carrier frequency to obtain a resonance frequency offset degree index; constructing a virtual grid impedance model based on the resonance frequency offset degree index to obtain equivalent network impedance parameters in the electromagnetic environment; performing correction calculation on the multi-dimensional electromagnetic interference characteristic model according to the equivalent network impedance parameters to obtain an enhanced interference characteristic model considering the influence of network impedance; designing a repetitive control compensation unit based on the enhanced interference characteristic model and connecting the repetitive control compensation unit to the target three-stage cascaded dynamic filtering network to obtain an enhanced filtering network with a repetitive controller; obtaining a capacitor voltage signal from the capacitor voltage sampling point of the double ball bearing fan and performing differential processing on the capacitor voltage signal to obtain a virtual capacitor current value; constructing a virtual grid-side current based on the virtual capacitor current value and the filtered Hall sensor signal to obtain a grid-side current equivalent signal independent of an additional current sensor; inputting the grid-side current equivalent signal into a passive constraint optimization algorithm for admittance reshaping calculation to obtain a damping optimization coefficient that satisfies the stability boundary condition; performing capacitor voltage feedforward active damping processing on the enhanced filtering network with a repetitive controller and adjusting the feedforward channel gain using the damping optimization coefficient to obtain an adaptive filtering control system with an extended stable region.

[0214] It should be noted that in this article, the terms "include", "comprise" or any other variant thereof are intended to cover non-exclusive inclusion, so that a process, apparatus, article or system including a series of elements not only includes those elements but also includes other elements not expressly listed, or also includes elements inherent to such process, apparatus, article or system. Without further limitation, an element defined by the statement "including a..." does not exclude the existence of additional identical elements in the process, apparatus, article or system including that element.

[0215] The above are only the preferred embodiments of the present invention, and do not limit the patent scope of the present invention accordingly. Any equivalent structure or equivalent process transformation made by using the content of the specification and drawings of the present invention, or directly or indirectly applied in other related technical fields, shall be similarly included in the patent protection scope of the present invention.

Claims

1. An intelligent cooling fan system with anti-electromagnetic interference based on the design of a dynamic filter circuit, characterized in that, Comprising: A dual - analysis module, which is used to perform dual - domain (time - domain and frequency - domain) analysis on the rotational speed feedback signal of a double - ball bearing fan, obtain a multi - dimensional electromagnetic interference characteristic model, and configure the filter network parameters of an initial three - stage cascaded dynamic filter network to obtain a target three - stage cascaded dynamic filter network; The dual - analysis module further includes: An acquisition unit, which is used to acquire the rotational speed feedback signal of the double - ball bearing fan within a preset rotational speed range, perform spectrum analysis on the rotational speed feedback signal to obtain an interference frequency distribution map; perform peak detection and energy statistics on the interference frequency distribution map to obtain the main interference frequencies and interference energy proportion coefficients within each frequency band, and perform mutation point detection and periodic analysis on the rotational speed feedback signal in the time domain to obtain interference waveform characteristics and time - domain distribution patterns; A parameter fitting unit, which is used to construct a four - dimensional electromagnetic interference correlation function based on the interference frequency distribution map, the main interference frequencies, the interference energy proportion coefficients, the interference waveform characteristics, and the time - domain distribution patterns; perform parameter fitting on the four - dimensional electromagnetic interference correlation function under different temperature conditions to obtain a multi - dimensional electromagnetic interference characteristic model including rotational speed - interference mapping, temperature - frequency offset, voltage - common - mode interference, and load - differential - mode interference relationships; A parameter configuration unit, which is used to configure the filter network parameters of the initial three - stage cascaded dynamic filter network based on the multi - dimensional electromagnetic interference characteristic model to obtain a target three - stage cascaded dynamic filter network; specifically, the parameter configuration unit is used to: configure the parameters of a π - type LC power filter circuit according to the voltage - common - mode interference relationship in the multi - dimensional electromagnetic interference characteristic model to obtain a first - stage filter circuit, and the first - stage filter circuit includes a common - mode choke and a low - ESR electrolytic capacitor; calculate the center frequency and bandwidth of a band - pass filter based on the rotational speed - interference mapping and temperature - frequency offset relationships in the multi - dimensional electromagnetic interference characteristic model to obtain a second - stage signal conditioning filter circuit; optimize and calculate the coefficient matrix of a digital filter based on the load - differential - mode interference relationship in the multi - dimensional electromagnetic interference characteristic model to obtain a third - stage digital filter circuit; cascade - connect the first - stage filter circuit, the second - stage signal conditioning filter circuit, and the third - stage digital filter circuit through an opto - isolator to obtain an initial three - stage cascaded dynamic filter network; perform a transfer - function test on the initial three - stage cascaded dynamic filter network under simulated electromagnetic interference conditions to obtain an actual filter characteristic curve and error parameters; adjust the parameters of the initial three - stage cascaded dynamic filter network according to the actual filter characteristic curve and the error parameters to obtain a target three - stage cascaded dynamic filter network; A signal purification module, which is used to acquire the first fan operation parameter of the double - ball bearing fan and input it into the target three - stage cascaded dynamic filter network for signal purification to obtain a filtered Hall sensor signal and a filtered power supply signal; A closed - loop control calculation module, which is used to perform closed - loop control calculation based on the filtered Hall sensor signal and the filtered power supply signal to obtain a non - linear PWM duty - cycle control quantity and phase distribution parameters; A phase difference control module, configured to perform phase difference control based on the non-linear PWM duty ratio control quantity and the phase distribution parameter, so as to obtain a multi-fan cooperative control strategy.

2. The anti-electromagnetic interference intelligent cooling fan system based on the dynamic filter circuit design according to claim 1, characterized in that, Specifically, the signal purification module is configured to: Collect first fan operation parameters from the Hall sensor, power supply monitoring port, temperature sensor and current sampling circuit of the double ball bearing fan; Input the power supply voltage waveform in the first fan operation parameters into the first-stage filtering circuit of the target three-stage cascaded dynamic filtering network to suppress common-mode interference, so as to obtain an initial power supply signal; Input the rotation speed pulse signal in the first fan operation parameters into the second-stage signal conditioning and filtering circuit of the target three-stage cascaded dynamic filtering network for band-pass filtering and automatic gain control, so as to obtain an initial Hall sensor signal; Perform temperature compensation calculation on the gain parameter of the second-stage signal conditioning and filtering circuit based on the ambient temperature value in the first fan operation parameters, so as to obtain a corrected signal gain coefficient; Input the initial power supply signal and the initial Hall sensor signal into the third-stage digital filtering circuit of the target three-stage cascaded dynamic filtering network for adaptive IIR filtering processing according to the corrected signal gain coefficient, and dynamically adjust the filter order according to the motor current value in the first fan operation parameters, so as to obtain a filtered Hall sensor signal and a filtered power supply signal.

3. The anti-electromagnetic interference intelligent cooling fan system based on the dynamic filter circuit design according to claim 1, characterized in that Specifically, the closed-loop control calculation module is configured to: Perform edge detection and period calculation on the filtered Hall sensor signal to obtain an actual rotation speed value and a rotation speed stability factor, and compare and calculate the actual rotation speed value with a preset rotation speed target value to obtain a rotation speed error value and a rotation speed change rate; Perform PID control parameter calculation based on the rotation speed error value, the rotation speed change rate and the rotation speed stability factor to obtain a linear control quantity; Input the linear control quantity into a piecewise linear mapping function for non-linear transformation processing to obtain a non-linear PWM duty ratio control quantity, and the piecewise linear mapping function adopts different slope values in different rotation speed intervals; Perform phase optimization calculation based on the number and physical layout information of the double ball bearing fans to obtain a phase distribution parameter.

4. The anti-electromagnetic interference intelligent cooling fan system based on the dynamic filter circuit design according to claim 1, characterized in that, Specifically, the phase difference control module is configured to: Send the non-linear PWM duty ratio control quantity and the phase distribution parameter to each slave controller in the distributed control network through the RS-485 bus to obtain a control frame data packet; Perform fan number self-adaptation calculation on the phase distribution parameter in the control frame data packet to obtain a phase difference between adjacent fans, and perform offset processing on the PWM signal generation timing of each slave controller based on the phase difference to obtain a multi-channel PWM drive signal; Perform real-time monitoring and analysis on the current waveforms of the multi-channel PWM drive signals to obtain total current ripple characteristics and electromagnetic interference degree data between each fan, and perform matching selection on the working modes of each fan based on the total current ripple characteristics and the electromagnetic interference degree data to obtain a working mode allocation scheme; Combining the working mode allocation scheme with the non-linear PWM duty cycle control quantity, dynamically adjusting the actual speed allocation of each fan according to the load factor and heat dissipation requirements of each fan, to obtain a multi-fan collaborative control strategy.

5. The anti-electromagnetic interference intelligent cooling fan system based on the dynamic filter circuit design according to claim 4, characterized in that, The intelligent heat dissipation fan system against electromagnetic interference based on the dynamic filter circuit design further includes: An execution module, configured to execute the multi-fan collaborative control strategy, and collect the operating parameters of the second fan from the distributed control network, where the operating parameters of the second fan include electrical parameters, mechanical parameters, and control parameters; A multi-dimensional determination module, configured to perform stability analysis and anomaly determination on the electrical parameters in the operating parameters of the second fan to obtain an electrical parameter health index, where the electrical parameters include power supply voltage stability, motor current waveform, power factor, and insulation resistance; perform noise spectrum and vibration characteristic analysis on the mechanical parameters in the operating parameters of the second fan to obtain a mechanical parameter health index, where the mechanical parameters include bearing noise, vibration amplitude, starting characteristics, and speed stability; perform response characteristic and anti-interference ability analysis on the control parameters in the operating parameters of the second fan to obtain a control parameter health index, where the control parameters include control accuracy, response time, anti-interference ability, and control margin; A response output module, configured to perform weighted fusion on the electrical parameter health index, the mechanical parameter health index, and the control parameter health index to obtain a multi-dimensional parameter health index, and perform hierarchical response processing based on the comparison result between the multi-dimensional parameter health index and a preset health threshold, and output a corresponding filter parameter recalibration instruction.

Images