Control method of multi-fan array

Abstract

The invention relates to the field of equipment control, in particular to a control method of a multi-fan array. The method comprises the steps that fan impeller angular displacement signals are collected through a displacement sensor array and converted into digital input signals, differential processing is conducted on the signals, a phase deviation value is calculated, a pulse width modulation coefficient is generated, PWM control parameters are synchronously transmitted through an EtherCAT bus to drive the direct-current brushless fan, a closed-loop feedback mechanism is established, and a master-slave type topological structure is adopted. The method also comprises the technical means of displacement sensor selection, bus setting, dynamic compensation algorithm, main control unit configuration, redundancy fault tolerance, topology identification, environmental parameter compensation, energy recovery, coding variable bit width and the like. According to the invention, the synchronization error lt of more than multiple fans is realized; the anti-interference capability, the time synchronization precision and the fault switching speed are improved, and the technical effects of self-adaptive load balancing, wind speed stable control, reduction of the overall power consumption of the system, self-adaptive adjustment of the data throughput and the like are achieved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent device control, and in particular to a control method for a multi-fan array. BACKGROUND

[0002] In many fields such as industrial production and electronic device heat dissipation, fans are widely used as common ventilation and heat dissipation equipment. With the progress of science and technology, there are more and more scenarios in which multiple fans work together to meet higher heat dissipation requirements or achieve specific ventilation functions. Effective control of the multi-fan array is a key factor for improving the performance and stability of the entire system. Its reasonable control not only relates to the normal operation of the equipment, but also has an important influence on energy utilization efficiency, equipment service life, etc. For example, in a large data center, a large number of servers generate heat that needs to be dissipated by a large number of fans, and the control of the multiple fans is directly related to the energy consumption and reliability of the data center; in an industrial workshop, multiple fans are also used for ventilation to ensure air quality and a suitable working environment for the equipment.

[0003] Traditionally, in order to control the on-off of several fans, the commonly used way is to connect multiple fans in parallel with a control box through connecting lines. Referring to Figure 1 This way can simply and directly achieve unified power supply and basic on-off control of multiple fans. The operator can relatively conveniently turn on or off all the fans connected with the control box. In addition, in some slightly more complex control scenarios, some simple circuit designs may be combined to achieve preliminary adjustment of the fan speed, but overall, these adjustment means are relatively limited and not accurate enough.

[0004] However, the traditional multi-fan control method has obvious defects. The parallel connection of multiple fans with the control box through connecting lines will result in too large current of the bus. The excessive current not only increases line loss and energy waste, but also causes the line to heat up seriously, which poses a safety hazard, and also puts higher requirements on the carrying capacity of the control box and related circuit elements, increasing equipment cost and maintenance difficulty. SUMMARY

[0005] The purpose of the present application is to overcome the above technical problems, and a control method for a multi-fan array is provided A control method for a multi-fan array, comprising the following steps: (1) Real-time acquisition of each fan impeller angular displacement signal by a displacement sensor array, and conversion of mechanical displacement to digital input signal DI by using a Gray code encoder; (2) In the distributed control module, the DI signal is subjected to differential processing and the phase deviation value Δθ is calculated to generate the pulse width modulation coefficient K _pwm = Δθ / (2π) x 100%; (3) The PWM control parameters of each node are synchronously transmitted to the corresponding fan driving unit through the EtherCAT bus, and a digital control signal DO is output to drive the direct-current brushless fan; (4) A closed-loop feedback mechanism is established: when the rotational speed deviation δ _n of any fan is detected to be greater than 5% of the rated value, a dynamic compensation algorithm is triggered to update the DO output, and the compensation amount ΔV=K _p ×δ _n + K _i ∫δ _n dt. (5) A master-slave topology structure is adopted, the master control unit verifies the data integrity of each slave node through CRC check code, and the synchronization error of more than one fan is less than 0.1 ms.

[0006] By adopting the above technical solutions, the fan impeller angular displacement signal can be collected in real time and converted into a digital input signal, the pulse width modulation coefficient is generated through differential processing and calculation, the PWM control parameters are synchronously transmitted to the fan driving unit to drive the direct-current brushless fan, the closed-loop feedback mechanism can dynamically compensate the rotational speed deviation, the master-slave topology structure can verify the data integrity of each slave node, and the synchronization error of more than one fan is less than 0.1 ms.

[0007] Preferably, the displacement sensor array in step (1) adopts a magnetic grid encoder, the measurement resolution reaches 0.01°, and the encoder output interface is configured with RS485 differential signal transmission with an anti-interference ability >60dB.

[0008] By adopting the above technical solutions, the displacement sensor array adopts a magnetic grid encoder to make the measurement resolution reach 0.01°, thereby improving the accuracy of the impeller angular displacement signal collection; the encoder output interface is configured with RS485 differential signal transmission to make the anti-interference ability >60dB, thereby ensuring the stability of the digital input signal transmission.

[0009] Preferably, the EtherCAT bus cycle in step (3) is set to 1ms, a distributed clock mechanism is adopted to realize time synchronization of each node, and the jitter compensation accuracy is <100ns.

[0010] By adopting the technical scheme, the EtherCAT bus cycle is set to 1ms, and the distributed clock mechanism is adopted to realize time synchronization of each node, so that the time synchronization of each node and the jitter compensation precision are less than 100ns, and the steps of cooperating with the displacement sensor array to collect the angular displacement signals of each fan impeller in real time, performing differential processing on the signals to calculate the phase deviation value and generate the pulse width modulation coefficient, transmitting the PWM control parameters to the corresponding fan driving unit through the bus to drive the direct-current brushless fan, establishing a closed-loop feedback mechanism, and verifying the data integrity of each slave node by adopting the master-slave topology structure can better realize the control of the multi-fan array and improve the precision and stability of the control.

[0011] Preferably, the dynamic compensation algorithm comprises a feedforward control link, and the compensation amount ΔV' is dynamically adjusted according to the feedback value P _real of the wind tunnel pressure sensor. _set _real _max .

[0012] By adopting the technical scheme, the angular displacement signals of each fan impeller are collected by the displacement sensor array and converted into digital input signals, the signals are differentially processed to calculate the phase deviation value and generate the pulse width modulation coefficient, the control parameters are synchronously transmitted through the EtherCAT bus to drive the direct-current brushless fan, the closed-loop feedback mechanism is established to trigger the compensation algorithm when the speed deviation exceeds 5% of the rated value, the master-slave topology structure is adopted to verify the data integrity to realize that the synchronization error of the multi-fan is less than 0.1ms, and the dynamic compensation algorithm comprises a feedforward control link to dynamically adjust the compensation amount according to the feedback value of the wind tunnel pressure sensor, thereby improving the precision and dynamic adaptability of the fan speed control.

[0013] Preferably, the master control unit integrates an FPGA chip to realize multi-channel parallel processing, and is configured with a double-redundancy CAN bus interface with a fault switching time of less than 50ms.

[0014] By adopting the technical scheme, the angular displacement signals of each fan impeller are collected by the displacement sensor array in real time and converted into digital input signals, the pulse width modulation coefficient is generated through the distributed control module, the control parameters are transmitted through the EtherCAT bus to drive the direct-current brushless fan, the closed-loop feedback mechanism is established to compensate the speed deviation, and the master-slave topology structure is adopted to realize that the synchronization error of the multi-fan is less than 0.1ms; on this basis, the master control unit integrates an FPGA chip to realize multi-channel parallel processing, can improve the data processing efficiency, is configured with a double-redundancy CAN bus interface with a fault switching time of less than 50ms, can ensure that the system is quickly recovered when a fault occurs, and guarantees the reliability and stability of data transmission.

[0015] ​​Preferably, in step (5), a redundant fault-tolerant mechanism is added, when any node communication interruption is detected, the backup data transmission path is automatically enabled, the backup path adopts a mixed structure of daisy chain and star topology, the optimal path is selected by a dynamic routing algorithm, the switching response time is ≤50 ms, and each fan node is configured with a double buffer area, the forward buffer area stores the current node DO output data, and the backward buffer area stores the upstream node DI input data, and the automatic retransmission when the data packet CRC check fails is realized.

[0016] By adopting the above technical scheme, the fan wheel angle displacement signals are collected in real time and converted into digital input signals, the signals are subjected to differential processing and calculation, the control parameters are synchronously transmitted to the corresponding fan driving units to drive the direct-current brushless fan, a closed-loop feedback mechanism is established to compensate the speed deviation, a master-slave type topology structure is adopted to realize multi-fan synchronous control, on this basis, a redundant fault-tolerant mechanism is added, the backup path can be automatically enabled when the node communication is interrupted, and the optimal path is selected by a dynamic routing algorithm, the system continuous operation is ensured by fast switching response, and the double buffer area can realize automatic retransmission when the data packet CRC check fails, thereby improving the data transmission reliability.

[0017] Preferably, a topology identification step is added before step (1), a topology detection pulse sequence is sent, a physical topology mapping table is constructed based on the fan node response delay Δtn=(n-1)×Tprop+Σ(i=1 to n)Tproc _i (wherein Tprop is a single node transmission delay, Tproc _i is the i-th node processing delay), the data transmission timing is dynamically adjusted, and adaptive load balancing of 1-multiple fan arrays is realized.

[0018] By adopting the above technical scheme, the fan wheel angle displacement signals are collected in real time by a displacement sensor array and converted into digital input signals, the signals are subjected to differential processing and calculation to obtain a pulse width modulation coefficient, the control parameters are synchronously transmitted by an EtherCAT bus to drive the direct-current brushless fan, a closed-loop feedback mechanism is established to trigger a dynamic compensation algorithm, a master-slave type topology structure is adopted to verify the data integrity of the slave node to realize multi-fan synchronization, on this basis, a topology identification step is added, a topology detection pulse sequence is sent, and a physical topology mapping table is constructed based on the fan node response delay, the data transmission timing can be dynamically adjusted, and adaptive load balancing of 1-multiple fan arrays is realized.

[0019] Preferably, in step (4), the environmental parameter compensation factor a = 0.8^(ΔT / 10) x log(1+Pamb / P0) (where ΔT is the environmental temperature change, and Pamb is the atmospheric pressure change) is introduced, the data is collected in real time by the temperature and pressure sensor integrated in the fan node, the PWM control coefficient K'pwm = Kpwm x a is dynamically corrected, and the wind speed control error is stabilized at ±2% in the environment range of -40℃ to 85℃.

[0020] By adopting the technical scheme, the displacement sensor array is used to collect the angular displacement signals of the fan impellers in real time and convert them into digital input signals, the signals are subjected to differential processing and calculation to obtain the pulse width modulation coefficient, the PWM control parameters are synchronously transmitted through the EtherCAT bus to drive the direct-current brushless fan, a closed-loop feedback mechanism is established to compensate the speed deviation, the master-slave topology structure is used to verify the data integrity to realize the multi-fan synchronous error control, and on this basis, the environmental parameter compensation factor is introduced in step (4), the data is collected in real time by the temperature and pressure sensor integrated in the fan node, the PWM control coefficient is dynamically corrected, and the wind speed control error can be stabilized at ±2% in the environment range of -40℃ to 85℃.

[0021] Preferably, in step (3), the energy recovery module is added, the braking energy is stored to the super capacitor through the reverse electromotive force capture circuit when the fan is decelerated, the energy storage efficiency is greater than or equal to 85%, and the energy scheduling protocol is established between adjacent fan nodes to realize the on-demand distribution of the energy in the array, and the overall power consumption of the system is reduced by 18%-22%.

[0022] By adopting the technical scheme, in the multi-fan array control, the energy recovery module is added, the braking energy can be stored to the super capacitor through the reverse electromotive force capture circuit when the fan is decelerated, the energy storage efficiency is high, the energy scheduling protocol can be established between adjacent fan nodes to realize the on-demand distribution of the energy in the array, and the overall power consumption of the system is effectively reduced.

[0023] Preferably, the displacement digital-to-analog encoding adopts a variable bit width mechanism, the 12-bit Gray code encoding is automatically switched to when the turbulence intensity of the wind tunnel test section is greater than or equal to 15%, the 8-bit encoding is adopted in the conventional working condition (turbulence intensity is less than 15%), the data throughput is adaptively adjusted in the range of 1.2-4.8Mbps, and the encoding switching response time is less than or equal to 10ms.

[0024] By adopting the technical scheme, the variable bit width mechanism is used for displacement digital-to-analog encoding, the 12-bit Gray code encoding is automatically switched to when the turbulence intensity of the wind tunnel test section is greater than or equal to 15%, the 8-bit encoding is adopted in the conventional working condition, the data throughput is adaptively adjusted in the range of 1.2-4.8Mbps, the encoding switching response time is less than or equal to 10ms, and the encoding flexibility and data transmission adaptability are improved.

[0025] In summary, the present application includes at least one of the following beneficial technical effects: 1. By means of signal acquisition by displacement sensor array, differential processing, EtherCAT bus transmission, etc., precise control and driving of the fan are realized, the problem of large bus current in traditional parallel mode is avoided, and line loss and energy waste are reduced; 2. A closed-loop feedback mechanism and a dynamic compensation algorithm are established, which can compensate in time according to the fan speed deviation and ensure stable operation of the fan; 3. The master-slave topology structure and CRC check are adopted, so that high-precision synchronous control of more than one fan can be realized, and the performance and stability of multi-fan cooperative work are improved. BRIEF DESCRIPTION OF DRAWINGS

[0026] Figure 1 is a connection relationship diagram of multiple fans and a control box in a traditional technology.

[0027] Figure 2 is a connection relationship diagram of multiple fans and a control box in an embodiment of the present application.

[0028] Figure 3 is a flowchart of a control method of a multi-fan array in an embodiment of the present application. DETAILED DESCRIPTION

[0029] The technical solutions in the embodiments of the present application will be described in detail below with reference to the accompanying drawings. Figures 2-3 The technical solutions in the embodiments of the present application will be described in detail below with reference to the accompanying drawings.

[0030] The present application mainly adopts displacement digital-to-analog encoding technology to control a multi-fan array, so as to achieve the effects of reducing bus current, precisely controlling the fan, and saving energy, etc. The present application will be described in further detail as follows.

[0031] Embodiment 1 The control method of the multi-fan array provided in the embodiment of the present application includes the following steps: (1) Real-time acquisition of each fan impeller angular displacement signal by a displacement sensor array, conversion of mechanical displacement to digital input signal DI by using Gray code encoder; (2) Differential processing of the DI signal and calculation of phase deviation value Δθ in the distributed control module, generation of pulse width modulation coefficient K _pwm = Δθ / (2π) × 100%; (3) Synchronous transmission of each node PWM control parameter to the corresponding fan driving unit through the EtherCAT bus, output of digital control signal DO to drive the direct-current brushless fan; (4) Establish a closed-loop feedback mechanism: when detecting the speed deviation δ of any fan _n >5% of the rated value, trigger the dynamic compensation algorithm to update the DO output, compensation ΔV = K _p × δ _n + K _i ∫δ _n dt; (5) Use master-slave topology structure, the master unit verifies the data integrity of each slave node through CRC check code, and realizes the synchronization error of more than one fan <0.1ms.

[0032] Correspondingly, the above steps are actually real-time acquisition of angular displacement signal, processing of signal to generate pulse width modulation coefficient, transmission of control parameter to drive motor, establishment of closed-loop feedback mechanism, and synchronization of fans using master-slave topology structure. Through the cooperation of these steps, the operation of multi-fan array can be accurately controlled, the problems such as excessive bus current caused by traditional parallel control mode can be reduced, and the stability and control accuracy of the system can be improved.

[0033] Specifically, the first step is to acquire the angular displacement signal of each fan impeller through the displacement sensor array. Here, Gray code encoder is used to convert mechanical displacement into digital input signal DI. The displacement sensor array uses magnetic grating encoder, which has a measurement resolution of 0.01°. The encoder output interface is configured with RS485 differential signal transmission, with an anti-interference ability of >60dB, which can effectively resist external interference and ensure the accuracy of signal transmission. Of course, in addition to the magnetic grating encoder, an optical encoder can also be used, which can also convert mechanical displacement into digital signal. However, the optical encoder may be relatively weak in dust resistance, but it can work well in a clean environment. Gray code encoder is a special binary code that can avoid the instantaneous errors that may occur during conversion of ordinary binary code, making the conversion of displacement to digital signal more reliable.

[0034] Next, in the distributed control module, the DI signal is differentially processed and the phase deviation value Δθ is calculated, and then the pulse width modulation coefficient K _pwm = Δθ / (2π)×100% is generated. The distributed control module can be composed of multiple small control chips, which are distributed at each node, so that the collected signals can be processed more quickly. Differential processing is a common signal processing method that can eliminate common-mode interference and highlight useful signal components, so as to more accurately calculate the phase deviation value.

[0035] After that, the PWM control parameters of each node are transmitted synchronously to the corresponding fan drive unit through the EtherCAT bus, and the digital control signal DO is output to drive the DC brushless fan. The EtherCAT bus cycle is set to 1ms, and the distributed clock mechanism is used to realize time synchronization of each node, with a jitter compensation accuracy of <100ns. The EtherCAT bus has high-speed data transmission capability, which can quickly transmit control parameters to each fan drive unit. The distributed clock mechanism can ensure the time synchronization of each node, making the operation of the entire system more coordinated. After receiving the control parameters, the fan drive unit accurately controls the speed of the DC brushless fan according to the digital control signal DO.

[0036] Then a closed-loop feedback mechanism is established. When the speed deviation δ _n of any fan is detected to be more than 5% of the rated value, the dynamic compensation algorithm is triggered to update the DO output, with a compensation amount ΔV = K _p ×δ _n + K _i ∫δ _n dt. Moreover, the dynamic compensation algorithm contains a feedforward control link, which dynamically adjusts the compensation amount ΔV' = ΔV × [1 + 0.2 × (P _real - P _set ) / P _real ] according to the feedback value P _max of the wind tunnel pressure sensor. The closed-loop feedback mechanism is like an intelligent regulation system that continuously monitors the speed of the fan and adjusts it in time once the speed deviation exceeds the set value. The dynamic compensation algorithm considers the current speed deviation and the integral of the deviation over a period of time, and combines the feedforward control link to more comprehensively compensate for the fan speed.

[0037] Finally, a master-slave topology structure is adopted, and the master control unit verifies the data integrity of each slave node through CRC checksum, achieving a synchronization error of <0.1ms for multiple fans. The master control unit integrates FPGA chips to realize multi-channel parallel processing and configures dual-redundancy CAN bus interfaces with a fault switching time of <50ms. In the master-slave topology structure, the master control unit is responsible for overall management, and the slave nodes perform specific tasks. The CRC checksum can effectively detect errors in data transmission and ensure data accuracy. In this way, a large number of fans can be synchronized to run with minimal error.

[0038] The implementation principle of the embodiment is: the embodiment can accurately control the operation of the multi-fan array by real-time acquisition of fan impeller angular displacement signals, a series of signal processing and transmission, combination of closed-loop feedback mechanism and master-slave topology structure. Compared with the traditional parallel control mode, the problem of excessive bus current is avoided, and energy waste and safety hazards are reduced. At the same time, the high-precision control can dynamically adjust the fan speed according to the actual situation, improve the stability and energy utilization efficiency of the system, and prolong the service life of the equipment.

[0039] Embodiment 2: The difference between the embodiment and the above-mentioned embodiment is that a plurality of optimizations are carried out on the basis of the original steps. A topology identification step is added before step (1), a physical topology mapping table is constructed based on the response time delay of each fan node Δtn=(n - 1)×Tprop + Σ(i = 1 to n)Tproc _i (where Tprop is the transmission time delay of a single node, and Tproc _i is the processing time delay of the i-th node), the data transmission time sequence is dynamically adjusted, and adaptive load balancing of 1-to-multiple fan arrays is realized. The topology identification step is like making a "map" for the system, which clearly shows the positions and states of each fan node. In this way, the time of data transmission can be reasonably allocated according to the actual situation, avoiding the situation where some nodes are congested with data transmission while some nodes are idle.

[0040] In step (4), an environmental parameter compensation factor α = 0.8^(ΔT / 10)×log(1 + Pamb / P0) (where ΔT is the environmental temperature change, and Pamb is the atmospheric pressure change) is introduced, the data collected by the temperature and pressure sensors integrated in the fan nodes are used to dynamically correct the PWM control coefficient K'pwm = Kpwm×α, so that the wind speed control error is stabilized at ±2% within the environmental range of -40°C~85°C. The environmental parameter compensation factor takes into account the influence of environmental temperature and atmospheric pressure on fan performance. By real-time acquisition of these parameters and correction of the control coefficient, the fan can operate stably under different environmental conditions.

[0041] In step (3), an energy recovery module is added. When the fan decelerates, the braking energy is stored in the super capacitor through the reverse electromotive force capture circuit, the energy storage efficiency is ≥85%, and an energy scheduling protocol is established between adjacent fan nodes to realize on-demand allocation of energy within the array, reducing the overall power consumption of the system by 18% - 22%. The energy recovery module is like a "small battery" that can recover and reuse the energy wasted when the fan decelerates. Through the energy scheduling protocol, the recovered energy can be reasonably allocated, greatly reducing the overall power consumption of the system.

[0042] The displacement digital code adopts a variable bit width mechanism, and automatically switches to 12-bit Gray code coding when it is detected that the turbulence intensity of the wind tunnel test section is greater than or equal to 15%, and 8-bit coding is adopted in a conventional working condition (turbulence intensity is less than 15%), the data throughput is adaptively adjusted in a range of 1.2-4.8 Mbps, and the coding switching response time is less than or equal to 10 ms. The variable bit width mechanism can automatically adjust the coding bit number according to different working conditions, while ensuring data accuracy, the data transmission efficiency is improved.

[0043] The implementation principle of the embodiment is that the embodiment further improves the performance of the multi-fan array control method by increasing a topology identification step, introducing an environmental parameter compensation factor, adding an energy recovery module, and adopting a variable bit width coding mechanism. The topology identification realizes adaptive load balancing, so that the system resources are more reasonably utilized; the environmental parameter compensation factor enables the fan to stably operate in different environments; the energy recovery module reduces the system power consumption and is more energy-saving and environmentally friendly; and the variable bit width coding mechanism improves the data transmission efficiency and enhances the adaptability and stability of the system.

[0044] Embodiment 3 The embodiment is different from the above-mentioned embodiments in that the fault tolerance and stability of the system are further strengthened. In step (5), a redundant fault tolerance mechanism is added, when it is detected that any node communication is interrupted, a backup data transmission path is automatically enabled, the backup path adopts a combination of a daisy chain and a star topology structure, the optimal path is selected through a dynamic routing algorithm, the switching response time is less than or equal to 50 ms, and each fan node is configured with double buffer areas, a forward buffer area stores the current node DO output data, and a backward buffer area stores the upstream node DI input data, so as to realize automatic retransmission when the data packet CRC check fails. The redundant fault tolerance mechanism is like an insurance for the system, when a node communication problem occurs, the system can quickly switch to a backup path to ensure normal operation. The combination of the daisy chain and the star topology structure combines the advantages of the two topology structures, and has good expansibility and high reliability. The setting of the double buffer areas can retransmit data in time when the data packet is wrong, and ensures the integrity of the data.

[0045] The implementation principle of the embodiment is that the embodiment improves the reliability and stability of the multi-fan array control system by adding a redundant fault tolerance mechanism. When facing sudden situations such as node communication interruption, the system can quickly respond and switch to a backup path, reducing the impact caused by the fault. The double buffer areas and the automatic retransmission mechanism ensure the accurate transmission of data, so that the entire system can stably work in a complex operating environment, and further improves the performance and availability of the system.

[0046] Embodiment 4 The embodiment provides a PWM speed regulation based multi-fan array control system, which is suitable for high reliability heat dissipation scenarios such as server cabinets, industrial equipment and the like. The system comprises a main controller and at most 100 DC 24V fan modules connected in parallel, each fan module satisfying the following technical parameters: working voltage range 18V-27.4V, rated current <1.2A (maximum locked-rotor current <3A), rated rotating speed 6500 RPM±10%, supporting PWM linear speed regulation (0%-100% duty cycle) and rotating speed / state feedback function.

[0047] System architecture and workflow: The main controller sends a unified PWM speed regulation signal (frequency 25kHz±5%, high level 2.8V-12V) to all fan modules through a single bus, and receives rotating speed pulses (2 pulses per rotation) and state signals (including locked-rotor, over-current and over-temperature alarms) output by OC gates of each module. After power-on, the system performs a self-checking process: if there is no digital communication signal within 2 seconds, the system automatically switches to an analog PWM speed regulation mode and runs at full speed with a 90% duty cycle by default. The main controller dynamically adjusts the PWM duty cycle, and the fan stabilizes to the target rotating speed (such as an upper limit of 7150 RPM) within 15 seconds. During the speed regulation process, the current waveform is monitored in real time (high-low ratio ≥70%, wide-narrow ratio ≥90%) to ensure that the electrical characteristics meet the requirements.

[0048] Protection mechanism and fault-tolerant control: When it is detected that the current of a certain fan module is ≥1.5A (over-current) or the rotating speed pulse is lost (locked-rotor), the main controller immediately cuts off the power supply of the module, and attempts to automatically restart after 5-15 seconds. If the restart fails for 3 consecutive times, the module is marked as a fault state, and the rotating speeds of the remaining fans are adjusted to compensate for the loss of air volume through a redundancy strategy. The system supports reverse connection protection (DC 24V reverse connection does not cause damage) and over-temperature protection (stopping rotation when the PCBA temperature is ≥95℃, and resuming when the temperature is below 80℃), and simultaneously protects the ESD interference of the PWM signal port through a TVS tube (meeting GB / T17626.2 Level 3).

[0049] Environmental adaptability and reliability verification: The mechanical structure of the fan module supports axial vertical, horizontal or inclined installation, the PCB adopts FR4 94V-0 flame-retardant material, and passes the following severe tests: Temperature cycle: after being stored for 24 hours at -40℃-85℃, 10 times of power-on and power-off impacts are performed at the minimum working voltage (18V), and then continuous operation is performed for 48 hours at extreme temperatures (-40℃ / 85℃), with a rotating speed deviation <15%; Damp heat aging: power-on operation is performed for 1000 hours in an environment of 85℃ / 85%RH, and the bearing has no wear (ASTM-D610 8-level standard); Mechanical impact: no structural damage after 100G half-sine wave impact (6ms pulse width, 3 times for each of XYZ axes); Lifetime verification: 50 samples run for 8568 hours at 85℃ (equivalent to L10 lifetime ≥ 70,000 hours at 40℃, IPC9591 standard).

[0050] Application examples: In a data center cabinet, the main controller dynamically adjusts the PWM duty cycle according to the temperature sensor data, so that all fans work cooperatively in the speed range of 50%–100%. If a fan triggers protection due to blocked rotation, the system automatically increases the speed of the remaining fans by 5% to maintain the total air volume, and sends a fault code to the upper computer through the RS485 bus.

[0051] This scheme realizes high-precision speed regulation and high-reliability operation of large-scale fan arrays through centralized control, distributed feedback, and multi-layer protection mechanisms, and is especially suitable for harsh working conditions such as offshore (D-class environment).

[0052] The above are preferred embodiments of the present application, and are not intended to limit the protection scope of the present application. Therefore, any equivalent changes made on the basis of the structure, shape, and principle of the present application should be covered within the protection scope of the present application.

Claims

1. A method for controlling a multi-fan array, characterized in that: The following steps are involved: (1) The angular displacement signal of each fan impeller is collected in real time through the displacement sensor array, and the mechanical displacement is converted into a digital input signal DI using a Gray code encoder; (2) In the distributed control module, the DI signal is differentially processed and the phase deviation value Δθ is calculated to generate the pulse width modulation coefficient K _pwm =Δθ / (2π)×100%; (3) The PWM control parameters of each node are synchronously transmitted to the corresponding fan drive unit through the EtherCAT bus, and the digital control signal DO is output to drive the DC brushless fan; (4) Establish a closed-loop feedback mechanism: When the speed deviation δ of any fan is detected _n When the rated value is >5%, the dynamic compensation algorithm is triggered to update the DO output, and the compensation amount ΔV=K _p ×δ _n + K _i ∫δ _n dt; (5) A master-slave topology is adopted, and the master control unit verifies the data integrity of each slave node through CRC check code, achieving a synchronization error of more than 0.1ms for multiple fans.

2. The method for controlling a multi-fan array according to claim 1, wherein: The displacement sensor array in step (1) adopts a magnetic grating encoder with a measurement resolution of 0.01°. The encoder output interface is configured with RS485 differential signal transmission and has an anti-interference capability of >60dB.

3. The control method of a multi-fan array according to claim 1, characterized in that: The EtherCAT bus cycle in step (3) is set to 1ms, and a distributed clock mechanism is used to achieve time synchronization of each node, with a jitter compensation accuracy of <100ns.

4. The method for controlling a multi-fan array according to claim 1, wherein: The dynamic compensation algorithm includes a feedforward control link, based on the wind tunnel pressure sensor feedback value P _real , dynamically adjust the compensation amount ΔV'=ΔV×[1+0.2×(P _set -P _real ) / P _max ].

5. The method for controlling a multi-fan array according to claim 1, wherein: The main control unit integrates an FPGA chip to achieve multi-channel parallel processing, is equipped with dual redundant CAN bus interfaces, and has a fault switching time of <50ms.

6. The method for controlling a multi-fan array according to claim 1, wherein: In step (5), a redundant fault-tolerant mechanism is added. When any node communication interruption is detected, the backup data transmission path is automatically enabled. The backup path adopts a hybrid structure of daisy chain and star topology, and the optimal path is selected through a dynamic routing algorithm. The switching response time is ≤50ms, and each fan node is configured with a double buffer area. The forward buffer area stores the current node DO output data, and the backward buffer area stores the upstream node DI input data, so as to realize automatic retransmission when the data packet CRC check fails.

7. The method for controlling a multi-fan array according to claim 1, wherein: Add a topology identification step before step (1), by sending a topology detection pulse sequence, based on the response delay of each fan node Δtn=(n-1)×Tprop+Σ(i=1 to n)Tproc _i (Tprop is the single-node transmission delay, Tproc _i Build a physical topology mapping table for the i-th node processing delay, dynamically adjust the data transmission timing, and achieve adaptive load balancing of one or more fan arrays.

8. The method for controlling a multi-fan array according to claim 1, wherein: In step (4), an environmental parameter compensation factor α = 0.8^(ΔT / 10) × log(1 + Pamb / P0) is introduced (where ΔT is the change in ambient temperature and Pamb is the change in atmospheric pressure). The temperature and pressure sensors integrated in the fan nodes collect data in real time, and the PWM control coefficient K'pwm = Kpwm × α is dynamically corrected to stabilize the wind speed control error within ±2% in the ambient temperature range of -40℃ ~ 85℃.

9. The method for controlling a multi-fan array according to claim 1, wherein: In step (3), an energy recovery module is added. When the fan decelerates, the braking energy is stored in the supercapacitor through the reverse electromotive force capture circuit, with an energy storage efficiency of ≥85%. An energy scheduling protocol is established between adjacent fan nodes to achieve on-demand energy distribution within the array, reducing the overall system power consumption by 18%-22%.

10. The method for controlling a multi-fan array according to claim 1, wherein: The displacement digital-to-analog encoding adopts a variable bit width mechanism. When the turbulence intensity in the wind tunnel test section is detected to be ≥15%, it automatically switches to 12-bit Gray code encoding. Under normal operating conditions (turbulence intensity <15%), 8-bit encoding is used. The data throughput is adaptively adjusted in the range of 1.2-4.8Mbps, and the code switching response time is ≤10ms.

Images