Server and edge management system thereof

Abstract

The invention discloses a server and an edge management system thereof, belongs to the field of servers, and considers that (1) BMC output information overtime can represent a BMC fault and (2) the local management efficiency can be improved through a voice control server, on one hand, the BMC can send preset management information to a control device after being started, and on the other hand, the BMC output information overtime can represent the BMC fault; on one hand, the control device can control the prompter to prompt a fault under the condition that the preset management information is overtime in receiving, on the other hand, the control device can convert a voice signal collected by the voice collection module into a control instruction and then send the control instruction to the CPU of the server, voice control over the server is achieved, BMC faults can be found in time locally in the server, and the service life of the server is prolonged. And the server can be conveniently and efficiently controlled through voice, so that the dependence of server management on the network is reduced, and the user experience is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of servers, in particular to a server and an edge management system thereof. BACKGROUND

[0002] In the era of big data, the importance of servers is increasingly reflected, so the management of servers is particularly important. However, there is a lack of a mature server management scheme in the related art. For the related management of servers, the server needs to transmit relevant information to a cloud device through a network so as to be processed by a computing device. The network dependence is high, and the local management of a computer room is difficult.

[0003] Therefore, how to provide a scheme for solving the above technical problems is a problem to be solved by those skilled in the art at present. SUMMARY

[0004] The present application aims to provide a server and an edge management system thereof. In the present application, on the one hand, the BMC can send preset management information to the control device after startup, and the control device can control the prompter to prompt a fault in the case of preset management information receiving timeout. On the other hand, the control device can send the control instruction converted from the voice signal collected by the voice collection module to the CPU of the server, so as to realize voice control of the server. The present application can not only discover the BMC fault in time locally, but also can conveniently and efficiently control the server through voice, reduces the dependence of server management on the network, and improves the user experience.

[0005] To solve the above technical problems, the present application provides an edge management system applied to a server, comprising:

[0006] A baseboard management controller configured to send preset management information of itself to a control device after startup;

[0007] A control device connected with the central processor of the server and the baseboard management controller, respectively, configured to control a prompter to prompt a baseboard management controller fault in the case of preset management information receiving timeout, and send the control instruction converted from the voice signal to the central processor of the server;

[0008] A prompter connected with the control device;

[0009] A voice collection module connected with the control device, configured to collect voice signals in the environment.

[0010] On the other hand, the control device comprises:

[0011] A microcontroller connected with the baseboard management controller and the prompter respectively, used for controlling the prompter to prompt the baseboard management controller failure in case that the preset management information is not received within a time limit, wherein the microcontroller is a microcontroller originally provided in the server;

[0012] An edge controller connected with the central processor of the server and the voice collection module respectively, used for sending the control instruction converted from the voice signal to the central processor of the server;

[0013] The prompter comprises a serial port screen.

[0014] On the other hand, the edge controller is further used for:

[0015] Based on the sensing data of the target sensor in the server, identifying whether the server has an abnormal condition through a preset abnormality identification strategy; and if the server has an abnormal condition, controlling the prompter to prompt the abnormal condition of the server.

[0016] On the other hand, the target sensor comprises a temperature sensor and a rotation speed sensor on a corresponding fan;

[0017] Based on the sensing data of the target sensor in the server, identifying whether the server has an abnormal condition through a preset abnormality identification strategy comprises:

[0018] Based on the output data of the temperature sensor and the rotation speed sensor, in case that the number of continuous data sampling points satisfying a preset abnormal condition exceeds a first preset threshold value, controlling the prompter to prompt a fan speed adjustment failure.

[0019] The preset abnormal condition is that a temperature sampling value of a current data sampling point satisfies a preset fan rotation speed adjustment rule, and the fan rotation speed does not change.

[0020] On the other hand, based on the output data of the temperature sensor and the rotation speed sensor, in case that the number of continuous data sampling points satisfying a preset abnormal condition exceeds a first preset threshold value, controlling the prompter to prompt a fan speed adjustment failure comprises:

[0021] Based on the output data of the temperature sensor and the rotation speed sensor, in case that the number of continuous data sampling points satisfying a preset abnormal condition exceeds a first preset threshold value, judging whether the output data of the target sensor collected by the baseboard management controller at a target sampling point is consistent with the output data of the target sensor collected by the edge controller at the target sampling point; and if the output data is consistent, controlling the prompter to prompt a fan speed adjustment failure.

[0022] On the other hand, the baseboard management controller is further used for:

[0023] When the server fails to start, the obtained failure identification code of the server is sent to the control device; and after the server successfully starts, the self-checking state of the server periodically generated is sent to the control device.

[0024] The control device is further configured to:

[0025] Control the prompter to prompt the specific failure content corresponding to the failure identification code, and control the prompter to prompt the self-checking state of the server received.

[0026] In another aspect, the voice collection module comprises:

[0027] The microphone array comprising at least three microphones is arranged in a ring shape, and the ring plane of the microphone array is parallel to the ground;

[0028] The central processor for converting the voice signal into the control instruction and then sending it to the server comprises:

[0029] After the sound source direction is determined through the microphone array, the voice signal is obtained from the sound source direction;

[0030] The control instruction extracted from the voice signal is sent to the central processor of the server.

[0031] In another aspect, after the sound source direction is determined through the microphone array, the voice signal is obtained from the sound source direction, which comprises:

[0032] When the microphone array is controlled to work in the cycle sound collection mode, if a voice signal with a decibel value between a first preset decibel and a second preset decibel is collected, then from the voice signals received within a preset time period, three voice signals with different receiving time and decibel values meeting a preset condition are selected;

[0033] According to the receiving time of the three voice signals and the position information of each microphone in the microphone array, the time difference of sound reaching each microphone is calculated;

[0034] Based on the time difference and the speed of sound propagation, the sound source direction is determined and the voice signal is obtained from the sound source direction;

[0035] The cycle sound collection mode is a mode in which a single microphone cyclically collects voice signals, and the preset condition is that the decibel value of the voice signal reaches a preset standard within the preset time period and the time interval between adjacent receiving times meets the set requirement.

[0036] In another aspect, the central processor for converting the voice signal into the control instruction and then sending it to the server further comprises:

[0037] When the microphone array is controlled to work in the cycle receiving mode, it is judged whether a voice signal with a second preset decibel value and a preset wake-up password is collected.

[0038] If the voice signal is collected, a control instruction is determined from a voice signal in a preset collection time length from a current time in the cycle receiving mode, and the control instruction is sent to a central processor of the server.

[0039] To solve the above technical problems, the application further provides a server comprising a central processor, and further comprising an edge management system as described above connected with the central processor.

[0040] Beneficial effects: The application provides an edge management system, which takes into account (1) BMC output information timeout can represent BMC failure, and (2) voice control server can improve local management efficiency, so in the application, on the one hand, the BMC can send preset management information to the control device after starting, and the control device can control the prompter to prompt failure in the case of preset management information receiving timeout, and on the other hand, the control device can send the control instruction converted from the voice signal collected by the voice collection module to the CPU of the server, to realize voice control of the server, which can not only discover BMC failure in time locally, but also can conveniently and efficiently control the server through voice, reduces the dependence of server management on the network, and improves user experience.

[0041] The application further provides a server, which has the same beneficial effects as the edge management system. BRIEF DESCRIPTION OF DRAWINGS

[0042] In order to more clearly illustrate the technical solutions in the embodiments of the application, the related art and the drawings needed in the embodiments will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the application, and other drawings can be obtained by those skilled in the art without creative effort.

[0043] Figure 1 A structural schematic diagram of an edge management system provided by the application;

[0044] Figure 2 A structural schematic diagram of another edge management system provided by the application;

[0045] Figure 3 A flowchart of a control instruction determination method provided by the application. DETAILED DESCRIPTION

[0046] The core of the present application is to provide a server and an edge management system thereof, in the present application, preset management information can be sent to a control device by a BMC after startup, the control device can control a prompter to prompt a fault in the case of preset management information receiving timeout, on the other hand, the control device can send the control instruction converted from the voice signal collected by the voice collection module to the CPU of the server, to realize voice control of the server, which can not only discover BMC fault in time locally on the server, but also can control the server conveniently and efficiently through voice, reduces the dependence of server management on network, and improves user experience.

[0047] To make the objects, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described below in connection with the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the protection scope of the present application.

[0048] Please refer to Figure 1 , Figure 1 The edge management system provided by the present application has a structural schematic diagram, which is applied to a server and comprises:

[0049] The substrate management controller 1 is used for sending preset management information of itself to the control device 2 after startup;

[0050] The control device 2 connected with the central processor of the server and the substrate management controller 1 respectively is used for controlling the prompter 3 to prompt the substrate management controller 1 fault in the case of preset management information receiving timeout, and sending the control instruction converted from the voice signal to the central processor of the server;

[0051] The prompter 3 connected with the control device 2;

[0052] The voice collection module 4 connected with the control device 2 is used for collecting voice signal in the environment.

[0053] Specifically, considering the technical problems in the background art above, and in combination with the consideration of (1) BMC output information timeout can represent BMC (Baseboard Management Controller) failure, (2) the local management efficiency can be improved by voice control server, in the embodiment of the present application, on the one hand, the preset management information of the BMC is sent to the control device 2 after starting, so that the control device 2 can monitor the "BMC sending action of preset management information" to realize the judgment of BMC failure, on the other hand, the voice signal of the voice acquisition module 4 can be determined to control instruction and sent to the CPU (Central Processing Unit) of the server, so as to realize the voice control of the server.

[0054] Specifically, based on the above consideration, the BMC in the edge management system in the embodiment of the present application can send its preset management information to the control device 2 after starting, and the voice acquisition module 4 can acquire the voice signal in the environment, and the control device 2 can control the prompter 3 to prompt the baseboard management controller 1 failure in the case of preset management information receiving timeout, and can send the control instruction converted from the voice signal to the central processing unit of the server, so as to realize the "BMC failure identification function" and "server voice control function", reduce the dependence of server management on network, and improve the user experience.

[0055] Wherein, the prompter 3 can be of various types, for example, it can be a serial port screen or a voice broadcaster, etc., which is not limited in the embodiment of the present application.

[0056] Specifically, the preset management information can be of various types, for example, it can include at least one of "network address, firmware version and manufacturer information of the baseboard management controller 1", etc., which is not limited in the embodiment of the present application.

[0057] The present application provides an edge management system, considering (1) BMC output information timeout can represent BMC failure, (2) the local management efficiency can be improved by voice control server, therefore, in the present application, on the one hand, the preset management information of the BMC can be sent to the control device after starting, and the control device can control the prompter to prompt the failure in the case of preset management information receiving timeout, on the other hand, the control device can send the control instruction converted from the voice signal collected by the voice acquisition module to the CPU of the server to realize the voice control of the server, which not only can discover the BMC failure in the server locally and timely, but also can control the server conveniently and efficiently through voice, reduce the dependence of server management on network, and improve the user experience.

[0058] On the basis of the above embodiment:

[0059] As an optional embodiment, the control device 2 comprises:

[0060] a microcontroller 21 connected with the baseboard management controller 1 and the prompter 3 respectively, for controlling the prompter 3 to prompt the baseboard management controller 1 failure in case of preset management information receiving timeout, wherein the microcontroller 21 is the original microcontroller 21 of the server;

[0061] an edge controller 22 connected with the central processor and the voice acquisition module 4 of the server respectively, for converting the voice signal into a control instruction and sending the control instruction to the central processor of the server;

[0062] the prompter 3 comprises a serial port screen.

[0063] Specifically, in order to better illustrate the embodiments of the present application, please refer to Figure 2 , Figure 2 the structure schematic diagram of another edge management system provided by the present application.

[0064] Specifically, considering that the original microcontroller 21 (MCU, Micro Controller Unit) of the server is used to perform the baseboard management controller 1 failure detection task, the system cost can be reduced; the edge controller 22 (EC, Edge Controller) is used to specially process the voice signal conversion, so as to improve the professionalism and efficiency of the system function; therefore, the control device 2 in the embodiments of the present application comprises the microcontroller 21 and the edge controller 22, the MCU is connected with the baseboard management controller 1 and the prompter 3 (serial port screen), and when the preset management information receiving timeout occurs, the MCU controls the serial port screen to display the baseboard management controller 1 failure. The EC is connected with the central processor and the voice acquisition module 4 of the server, and converts the acquired voice signal into a control instruction and sends the control instruction to the central processor.

[0065] The microcontroller 21 and the serial port screen can be connected through the serial port of the microcontroller 21, which reduces the difficulty of scheme design and has low cost, and the serial port can be of various types, such as UART (Universal Asynchronous Receiver / Transmitter, Universal Asynchronous Receiver / Transmitter), etc., which is not limited in the embodiments of the present application.

[0066] The serial port screen can be installed on the right upper side of the server ear plate to avoid blocking the heat dissipation port and can be externally provided with a transparent ABS (Acrylonitrile Butadiene Styrene) protective shell. The serial port of the serial port screen can reuse the cable channel of the idle VGA (Video Graphics Array) interface of the server to reduce the internal wiring complexity.

[0067] Of course, in addition to this specific configuration, the control device 2 can also have other specific configurations, which are not limited herein.

[0068] As an optional embodiment, the edge controller 22 is further configured to:

[0069] Based on the sensing data of the target sensor in the server, whether the server has an abnormal condition is identified through a preset abnormality identification strategy; if the server has an abnormal condition, the prompter 3 is controlled to prompt the abnormal condition of the server.

[0070] Specifically, considering that the edge controller 22 can assist in judging the abnormal state of the server based on the related sensing data in the server, so as to take timely measures to ensure the stable operation of the server, therefore, the edge controller 22 in the embodiment of the application can also identify whether the server has an abnormal condition based on the sensing data of the target sensor in the server through a preset abnormality identification strategy; if the server has an abnormal condition, the prompter 3 is controlled to prompt the abnormal condition of the server.

[0071] The target sensor can be of various types, such as a temperature sensor, a fan speed sensor, etc., and the corresponding abnormal condition can correspond to the target sensor, such as temperature being too high, fan failure, etc., which are not limited herein.

[0072] As an optional embodiment, the target sensor includes a temperature sensor and a speed sensor on a corresponding fan.

[0073] Based on the sensing data of the target sensor in the server, whether the server has an abnormal condition is identified through a preset abnormality identification strategy, which includes:

[0074] Based on the output data of the temperature sensor and the speed sensor, if the number of consecutive data sampling points that meet the preset abnormal condition exceeds a first preset threshold, the prompter 3 is controlled to prompt a fan speed adjustment failure.

[0075] The preset abnormal condition is that the temperature sampling value of the current data sampling point meets a preset fan speed adjustment rule, and the fan speed does not change.

[0076] Specifically, considering that the fan speed regulation failure will affect the server heat dissipation, cause the server to run abnormally, and through the data monitoring of the temperature sensor (TS, Temperature Sensor) and the rotation speed sensor (RSS, Rotation Speed Sensor), the fan speed regulation failure can be found in time in theory, and the normal work of the server heat dissipation system is ensured; therefore, in the embodiment of the application, the temperature sensor and the rotation speed sensor collect the temperature and fan rotation speed data of the server respectively. When the temperature sampling value at the current data sampling point meets the preset fan rotation speed adjustment rule, the fan rotation speed does not change (that is, the preset abnormal condition is met), and the number of continuous occurrence of the data sampling points meeting the condition exceeds the first preset threshold, the edge controller 22 controls the prompter 3 to prompt the fan speed regulation failure.

[0077] In the above embodiment, the preset fan rotation speed adjustment rule of the temperature sensor in the server is: when the temperature is greater than or equal to 60 DEG C, the fan rotation speed should be increased from 1000 revolutions per minute to 1500 revolutions per minute; and the first preset threshold is 3 times. The specific example is as follows:

[0078] The data sampling process is as follows: sampling point 1: the temperature sensor collects 62 DEG C (meets the adjustment rule of greater than or equal to 60 DEG C), but the rotation speed sensor feedbacks that the fan rotation speed is still 1000 revolutions per minute (does not increase according to the rule), at this time, the preset abnormal condition is met. Sampling point 2: the temperature continues to be 63 DEG C, and the fan rotation speed is still 1000 revolutions per minute, the abnormal condition is met again. Sampling point 3: the temperature is 61 DEG C, and the fan rotation speed does not change, the abnormal condition is met continuously for 3 times (exceeds the first preset threshold of 3 times).

[0079] The fault prompt triggering is as follows: the edge controller 22 detects that the continuous 3 sampling points all meet the abnormal condition of "temperature meeting the standard but rotation speed not adjusting", immediately controls the prompter 3 (such as a serial port screen) to display "fan speed regulation failure", and may be accompanied by light flashing or buzzer alarm, prompting the operation and maintenance personnel that the fan speed control mechanism is abnormal, and whether the fan drive circuit or the rotation speed sensor is faulty needs to be checked.

[0080] As an optional embodiment, based on the output data of the temperature sensor and the rotation speed sensor, in the case that the number of continuous occurrence of the data sampling points meeting the preset abnormal condition exceeds the first preset threshold, the control of the prompter 3 to prompt the fan speed regulation failure includes:

[0081] Based on the output data of the temperature sensor and the rotation speed sensor, in the case that the number of continuous occurrence of the data sampling points meeting the preset abnormal condition exceeds the first preset threshold, it is judged whether the output data of the target sensor collected by the baseboard management controller 1 at the target sampling point is consistent with the output data of the target sensor collected by itself at the target sampling point; if consistent, the control of the prompter 3 to prompt the fan speed regulation failure.

[0082] It is to be noted that, even in the case that the number of data sampling points meeting the preset abnormal condition continuously appears exceeds the first preset threshold, the result of the fan speed control failure judged by the edge controller 22 can also be wrong due to the reason of "sensing data acquisition error", and the consistency comparison with the sensing data collected by the BMC can improve the reliability of the result of the fan speed control failure judged by the edge controller 22, therefore, in the embodiment of the present application, the output data of the temperature sensor and the speed sensor can be used to judge whether the output data of the target sensor collected by the baseboard management controller 1 at the target sampling point is consistent with the output data of the target sensor collected by itself at the target sampling point in the case that the number of data sampling points meeting the preset abnormal condition continuously appears exceeds the first preset threshold; if consistent, the control prompter 3 is prompted to prompt the fan speed control failure; thereby the reliability of the failure judgment can be improved, the false alarm of the failure can be avoided, and the user experience can be further improved.

[0083] As an optional embodiment, the microcontroller 21 is further used to:

[0084] control the prompter 3 to prompt the received preset management information;

[0085] The preset management information includes at least one of the following:

[0086] The network address, firmware version and manufacturer information of the baseboard management controller 1.

[0087] Specifically, considering that the microcontroller 21 controls the serial port screen to prompt the received preset management information, the staff near the server can know the preset management information in time and manage the server, therefore, the microcontroller 21 in the embodiment of the present application can also control the prompter 3 to prompt the received preset management information.

[0088] As an optional embodiment, the baseboard management controller is further used to:

[0089] When the server fails to start, the obtained failure identification code of the server is sent to the control device 2; after the server successfully starts, the periodic self-test state of the server is sent to the control device 2;

[0090] The control device 2 is further used to:

[0091] control the prompter 3 to prompt the specific failure content corresponding to the failure identification code, and control the prompter 3 to prompt the received self-test state of the whole machine.

[0092] Specifically, considering that the BMC can obtain the server fault identification code when the server fails to start, and the specific fault content can be determined through the fault identification code, which can guide the staff to quickly locate the fault, therefore, the BMC in the embodiment of the present application can send the obtained server fault identification code to the control device 2 when the server fails to start, and the control device 2 can control the prompter 3 to prompt the specific fault content corresponding to the fault identification code.

[0093] In addition, considering that in the case of successful server startup, the periodic self-test state of the whole machine can guide the staff to comprehensively and quickly understand the server state, therefore, the BMC in the embodiment of the present application can send the periodic self-test state of the whole machine generated by the server to the control device 2 after the server starts successfully, and the control device 2 controls the prompter 3 to prompt the received self-test state of the whole machine.

[0094] Wherein, referring to Figure 2 The edge management system in the embodiment of the present application can also include a network communication module (such as a wireless network), and the edge controller 22 can send various prompt information and alarm information to the network terminal (such as the mobile phone of the operation and maintenance personnel) through the network communication module.

[0095] As an optional embodiment, the voice acquisition module 4 includes:

[0096] The ring-shaped arrangement includes a microphone array including at least three microphones, and the ring-shaped plane of the microphone array is parallel to the ground;

[0097] The voice signal is converted into a control instruction and then sent to the central processor of the server, including:

[0098] After the sound source direction is determined through the microphone array, the voice signal is obtained from the sound source direction;

[0099] The control instruction extracted from the voice signal is sent to the central processor of the server.

[0100] Specifically, considering that the sound source direction can be determined through the ring-shaped arrangement of the microphone array, so that the voice signal can be more accurately obtained and the control instruction can be extracted, which is beneficial to improve the precision of server voice control and further improve the user experience, therefore, the voice acquisition module 4 in the embodiment of the present application includes: a ring-shaped arrangement including a microphone array including at least three microphones, and the ring-shaped plane of the microphone array is parallel to the ground; the edge controller 22 can determine the sound source direction through the microphone array, and then obtain the voice signal from the sound source direction; and the control instruction extracted from the voice signal is sent to the central processor of the server.

[0101] The number of microphones included in the microphone array can be autonomously set, for example, 5 (equidistant annular arrangement), and the embodiments of the present application do not limit this.

[0102] As an optional embodiment, after the sound source direction is determined by the microphone array, the voice signal is obtained from the sound source direction, which includes:

[0103] When the microphone array is controlled to work in the cycle sound collection mode, if the voice signal with the decibel value between the first preset decibel and the second preset decibel is collected, three voice signals with different receiving time and decibel value meeting the preset condition are selected from the voice signals received within the preset time length;

[0104] According to the receiving time of the three voice signals and the position information of each microphone in the microphone array, the time difference of the sound reaching each microphone is calculated;

[0105] Based on the time difference and the speed of sound propagation, the sound source direction is determined and the voice signal is obtained from the sound source direction;

[0106] The cycle sound collection mode is a mode in which a single microphone cyclically collects voice signals, and the preset condition is that the decibel value of the voice signal reaches the preset standard within the preset time length and the time interval between adjacent receiving time meets the set requirement.

[0107] Specifically, considering that the sound source direction can be determined when a relatively weak sound (such as footsteps, cough or clothes rubbing sound) is collected, so that the subsequent real stronger voice signal (including the control password spoken by the user) can be more accurately received, therefore, in the embodiments of the present application, when the microphone array is controlled to work in the cycle sound collection mode (lower power consumption), if the voice signal with the decibel value between the first preset decibel and the second preset decibel is collected, three voice signals with different receiving time and decibel value meeting the preset condition are selected from the voice signals received within the preset time length, then according to the receiving time of the three voice signals and the position information of each microphone in the microphone array, the time difference of the sound reaching each microphone is calculated; Based on the time difference and the speed of sound propagation, the sound source direction is determined and the voice signal is obtained from the sound source direction; so that the user's sound source direction can be determined through the weak sound before the user really speaks the control password, and the voice signal is obtained from the sound source direction, which improves the accuracy of the voice control of the server.

[0108] The voice signal obtained from the sound source direction can include: controlling the microphones in the sound source direction to simultaneously obtain the voice signal (centralized sound collection mode).

[0109] As an optional embodiment, after the voice signal is converted into a control instruction and sent to the central processor of the server, it further includes:

[0110] In the case that the microphone array works in the cycle receiving mode, it is judged whether the voice signal with the second preset decibel and the preset wake-up password is collected;

[0111] If the voice signal with the second preset decibel and the preset wake-up password is collected, the control instruction is determined from the voice signal in the preset collection time length from the current time in the cycle receiving mode, and the control instruction is sent to the central processor of the server.

[0112] Specifically, in order to better illustrate the embodiments of the present application, please refer to Figure 3 , Figure 3 The flowchart of the control instruction determination method provided by the present application first controls the microphone array to work in the cycle receiving mode to collect the voice signal, then judges whether the voice signal reaches the second preset decibel, if the voice signal reaches the second preset decibel, it is judged that the voice signal is the preset wake-up password, if the voice signal is the preset wake-up password, the control instruction is determined from the voice signal in the preset collection time length from the current time in the cycle receiving mode, and the control instruction is sent to the central processor of the server, if the voice signal does not reach the second preset decibel, it is judged that the voice signal decibel is greater than the first preset decibel, if the voice signal decibel is greater than the first preset decibel, the sound source direction is determined first, and then the voice signal is obtained from the sound source direction, that is, "from the voice signal received in the preset time length, three voice signals with different receiving time and decibel value meeting the preset condition are selected, the time difference of the sound reaching each microphone is calculated according to the receiving time of the three voice signals and the position information of each microphone in the microphone array, the sound source direction is determined based on the time difference and the sound propagation speed, and the voice signal is obtained from the sound source direction", and finally the control instruction extracted from the voice signal is sent to the central processor of the server.

[0113] Specifically, considering that the wake-up password detection can avoid system false triggering, improve the safety and reliability of the system, and ensure that the voice instruction is processed only after the user issues a specific wake-up password, in the case that the decibel value of the voice signal reaches the second preset decibel and the voice content is the preset wake-up password, the control instruction is determined from the voice signal in the preset collection time length from the current time in the cycle receiving mode, and the control instruction is sent to the central processor of the server; that is, in the case that the user wakes up by the preset wake-up password with strong decibel, the cycle receiving mode with low power consumption is continued to "collect the voice signal in the preset collection time length from the current time" and determine the control instruction.

[0114] The preset wake-up password can be set autonomously, for example, it can include "wake-up server" and the like, which is not limited in the embodiments of the present application.

[0115] In addition, as an optional embodiment, the voice collection module 4 further comprises:

[0116] A noise suppression unit connected with the microphone array, for performing real-time noise reduction processing on the collected voice signal through an adaptive filtering algorithm, the adaptive filtering algorithm being trained based on a deep learning model and capable of identifying the type of environmental noise (including fan noise, current noise, background human voice) and dynamically adjusting the filtering parameters.

[0117] Specifically, considering that the noise in the environment (such as fan noise, current noise, and background human voice) will affect the quality of the voice signal, the noise interference is reduced through the noise suppression unit to improve the clarity and accuracy of the voice signal; therefore, the voice collection module 4 in the embodiment of the present application can comprise a noise suppression unit connected with the microphone array, for performing real-time noise reduction processing on the collected voice signal through an adaptive filtering algorithm, wherein the adaptive filtering algorithm is trained based on a deep learning model and capable of identifying the type of environmental noise (including fan noise, current noise, and background human voice) and dynamically adjusting the filtering parameters.

[0118] Among them, the noise suppression unit (NSU, Noise Suppression Unit) in the voice collection module 4 is connected with the microphone array, and can use an adaptive filtering algorithm (AFA, Adaptive Filtering Algorithm) trained based on a deep learning model (DLM, Deep Learning Model) to perform real-time noise reduction processing on the collected voice signal. The algorithm can identify the type of environmental noise and dynamically adjust the filtering parameters to optimize the quality of the voice signal.

[0119] Specifically, one specific embodiment of the voice collection module 4 comprises: a hardware part: a microphone array composed of 4 microphones arranged in a ring shape, with the ring plane parallel to the ground and installed on the top of the server cabinet. A noise suppression unit integrated on the circuit board of the edge controller 22 of the control device 2, connected with the microphone array through an SPI (Serial Peripheral Interface, Serial Peripheral Interface) bus. Software part: deep learning model (DLM): using a convolutional neural network (CNN, Convolutional Neural Network) architecture, pre-trained on a dataset containing 100,000 sets of computer room noise samples (fan noise, current noise, and background human voice), and the model parameters are stored in the Flash (Flash) memory of the NSU. Adaptive filtering algorithm (AFA): based on the trained DLM to identify the type of noise in real time and adjust the filtering parameters.

[0120] Among them, the working process of the voice collection module 4 comprises:

[0121] Real-time noise identification:

[0122] The microphone array collects voice signals (such as the "start server" instruction issued by the user) and noise signals in the machine room environment and transmits them to the NSU at a sampling rate of 16 kHz and a quantization accuracy of 16 bits.

[0123] The NSU converts the input time-domain signal into a frequency-domain feature (such as Mel-frequency cepstral coefficients, MFCC) and inputs it into the DLM. The DLM processes through convolution layers, pooling layers, and fully connected layers and outputs the probability distribution of the current noise type (such as a fan noise probability of 80%, an electrical noise probability of 15%, and a background human voice probability of 5%) to determine that the main noise type is fan noise.

[0124] Dynamic adjustment of filter parameters:

[0125] According to the identified fan noise type, the AFA retrieves the corresponding initial filter parameters from the preset parameter library: the band-stop filter center frequency is set to the main harmonic frequency of the fan (such as 300 Hz, 600 Hz, and 900 Hz), the bandwidth is set to 50 Hz, and the attenuation coefficient is set to -25 dB. At the same time, the AFA updates the filter parameters every 100 ms by monitoring the real-time changes of the noise signal (such as the frequency shift of the fan noise caused by the change in fan speed): if the fan speed is detected to be rising, the center frequency is automatically adjusted to 350 Hz, 650 Hz, and 950 Hz to ensure accurate suppression of the current noise.

[0126] Voice signal optimization processing:

[0127] The NSU uses the adjusted filter parameters to perform real-time filtering on the input signal: through the band-stop filter, the energy of the fan noise frequency band is attenuated, and at the same time, the adaptive gain control (AGC) is used to improve the volume of the voice signal (such as the "start server" instruction), so that the signal-to-noise ratio (SNR) is improved from the initial 10 dB to more than 25 dB. The clean voice signal after processing is converted into a control instruction by the edge controller 22 and sent to the server central processor.

[0128] The application also provides a server comprising a central processor and an edge management system as in the preceding embodiments connected to the central processor.

[0129] Embodiments of the present application are described herein with reference to the accompanying drawings, of which: various embodiments are described with progressive

[0130] The above description of disclosed embodiments is intended to be illustrative and not restrictive. Many embodiments of the application will be apparent to those of skill in the art upon reviewing the above description. The scope of the application should, therefore, be determined not with reference to the above description, but instead should be given with reference to the appended claims, along with their full scope of equivalents.

Claims

1. An edge management system, characterized in that: Applicable to servers, including: The baseboard management controller is used to send its own preset management information to the control device after startup; A control device connected to the central processing unit of the server and the baseboard management controller, respectively, is used to control the prompter to prompt the baseboard management controller failure when the preset management information reception times out; convert the voice signal into a control command and send it to the central processing unit of the server; a prompter connected to the control device; The voice collection module connected to the control device is used to collect voice signals in the environment.

2. The edge management system according to claim 1, characterized in that: The control device comprises: A microcontroller connected to the baseboard management controller and the prompter, respectively, for controlling the prompter to prompt a baseboard management controller failure when the preset management information reception times out, wherein the microcontroller is an original microcontroller of the server; An edge controller connected to the central processing unit of the server and the voice acquisition module, respectively, for converting the voice signal into a control instruction and sending the result to the central processing unit of the server; The prompter includes a serial port screen.

3. The edge management system according to claim 2, characterized in that: The edge controller is also used to: Based on the sensor data of the target sensor in the server, a preset abnormality identification strategy is used to identify whether the server has an abnormal condition; if an abnormal condition exists, the control prompter prompts the server of the abnormal condition.

4. The edge management system according to claim 3, characterized in that: The target sensor includes a temperature sensor and a speed sensor on its corresponding fan; Based on the sensor data of the target sensor in the server, the preset anomaly identification strategy is used to identify whether the server has an abnormal condition, including: Based on the output data of the temperature sensor and the speed sensor, when the number of data sampling points that meet the preset abnormal condition that appear continuously exceeds a first preset threshold, the control prompter prompts a fan speed regulation failure; The preset abnormal condition is: the temperature sampling value of the current data sampling point meets the preset fan speed adjustment rule, and the fan speed does not change.

5. The edge management system according to claim 4, characterized in that: The controlling the prompter to prompt a fan speed regulation fault based on the output data of the temperature sensor and the speed sensor when the number of consecutive data sampling points meeting the preset abnormal condition exceeds a first preset threshold includes: Based on the output data of the temperature sensor and the speed sensor, when the number of consecutive occurrences of data sampling points that meet the preset abnormal conditions exceeds a first preset threshold, it is determined whether the output data of the target sensor collected by the baseboard management controller at the target sampling point is consistent with the output data of the target sensor collected by the baseboard management controller at the target sampling point; if they are consistent, the control prompter prompts a fan speed regulation failure.

6. The edge management system according to claim 1, characterized in that: The baseboard management controller is further configured to: When the server fails to start up, the acquired fault identification code of the server is sent to the control device; after the server starts up successfully, the self-check status of the server generated periodically is sent to the control device; The control device is also used for: The control prompter prompts the specific fault content corresponding to the fault identification code, and the control prompter prompts the received whole machine self-test status.

7. The edge management system according to any one of claims 1 to 6, characterized in that: The voice acquisition module includes: a microphone array comprising at least three microphones arranged in a ring, wherein the ring plane of the microphone array is parallel to the ground; The central processing unit that converts the voice signal into a control instruction and sends it to the server includes: After determining the direction of the sound source through the microphone array, acquiring a voice signal from the direction of the sound source; The control instruction extracted from the voice signal is sent to the central processing unit of the server.

8. The edge management system according to claim 7, characterized in that: After the direction of the sound source is determined by the microphone array, acquiring a voice signal from the direction of the sound source includes: When the microphone array is controlled to operate in a cyclic sound collection mode, if a voice signal with a decibel value between a first preset decibel and a second preset decibel is collected, three voice signals with decibel values ​​meeting a preset condition at different reception times are selected from the voice signals received within a preset time period; Calculate the time difference between the sound reaching each microphone based on the reception time of the three voice signals and the position information of each microphone in the microphone array; Based on the time difference and the sound propagation speed, determining the direction of the sound source and obtaining a voice signal from the direction of the sound source; Among them, the cyclic sound collection mode is a mode in which a single microphone cyclically collects voice signals, and the preset condition is that the decibel value of the voice signal reaches a preset standard within the preset time length and the time interval between adjacent receiving moments meets the set requirements.

9. The edge management system according to claim 7, characterized in that: The central processor that converts the voice signal into a control instruction and sends it to the server also includes: When the microphone array is controlled to operate in a cyclic sound collection mode, determining whether a voice signal having a decibel value reaching a second preset decibel and a voice content being a preset wake-up command is collected; If collected, then in the cyclic sound collection mode, a control instruction is determined from the voice signal within the preset collection time from the current moment, and the control instruction is sent to the central processing unit of the server.

10. A server, characterized in that: The system comprises a central processing unit and an edge management system according to any one of claims 1 to 9 connected to the central processing unit.

Images