CN116594442A

CN116594442A - Control method and device of temperature regulation module and computing equipment

Info

Publication number: CN116594442A
Application number: CN202310409739.7A
Authority: CN
Inventors: 林良艺
Original assignee: XFusion Digital Technologies Co Ltd
Current assignee: XFusion Digital Technologies Co Ltd
Priority date: 2023-04-14
Filing date: 2023-04-14
Publication date: 2023-08-15

Abstract

The embodiment of the application discloses a control method, a device and a computing device of a temperature regulation module, wherein the method comprises the steps of determining a temperature control error and a temperature index parameter according to the temperatures of a plurality of temperature control points in the computing device; if the temperature index parameter meets the adjustment condition, determining a target control parameter according to the temperature control error and a control model; and controlling the operation of the temperature regulating module based on the target control parameter. In this way, temperature control can be flexibly achieved.

Description

Control method and device of temperature regulation module and computing equipment

Technical Field

The present application relates to the field of computing devices, and in particular, to a method and an apparatus for controlling a temperature adjustment module, and a computing device.

Background

The server is used as a high-performance computer and can provide various functions such as data analysis, publishing of multimedia data, information interaction, storage and the like for users through a network, and along with the increase of the functions of the server, the performance requirements on the server are higher and higher, and CPU, GPU, IPU and other diversified processing chips are fused into one server, so that the server becomes a multi-heat-source, high-dynamic and high-concurrency heating system. And along with the modularized design, the server can be compatible with various peripheral equipment and various functional modules for mixed use, and the heating situation is more complicated and various.

The existing heat dissipation schemes comprise air cooling heat dissipation, water cooling heat dissipation and the like, the heat dissipation schemes are used for conducting heat dissipation adjustment based on the temperature inside the server, the temperature control mode is single, and the heat dissipation effect is poor for the server with more heating sources.

Disclosure of Invention

The embodiment of the application provides a control method and device of a temperature regulation module and computing equipment, which can flexibly regulate and control equipment temperature.

In a first aspect, an embodiment of the present application provides a method for controlling a temperature adjustment module, including:

determining a temperature control error and a temperature index parameter according to the temperatures of a plurality of temperature control points in the computing equipment; if the temperature index parameter meets the adjustment condition, determining a target control parameter according to the temperature control error and a control model; and controlling the operation of the temperature regulating module based on the target control parameter.

In the technical scheme, whether the self-adaptive adjustment of parameters such as PID is carried out is judged through the temperature index parameters related to the acquired temperature, and under the condition that the temperature index parameters meet the adjustment conditions of the self-adaptive adjustment, the temperature control errors and the control model can be combined to generate control parameters, so that the temperature adjusting module such as a fan is controlled to work through the control parameters, the temperature of the computing equipment can be flexibly and accurately controlled from the measured temperature, the dynamic response to the temperature control is enhanced, and the temperature control can be carried out in a manner of adapting to faster temperature change.

In one implementation manner of this embodiment, if the temperature index parameter meets the adjustment condition, determining the target control parameter according to the temperature control error and the control model includes: if the temperature index parameter meets the adjustment condition, iterating the process control parameter at least once according to the temperature control error and the control model; wherein the process control parameter is configured to control the temperature adjustment module to operate for a preset duration; and if the temperature regulation module works based on the latest process control parameter so that the temperature index parameter does not meet the regulation condition, determining the latest process control parameter as the target control parameter.

In the implementation mode, a proper control parameter is finally determined through iterative processing, so that the control of the working temperature of the computing equipment such as a server is realized, the temperature control in various scenes can be effectively adapted, the automation and the intelligent degree of the temperature control are improved, and the more accurate heat dissipation control capability of the computing equipment under the conditions of working environment, parts, application changes and the like is realized.

In one implementation of this embodiment, after controlling the operation of the temperature adjustment module based on the target control parameter, the method further includes: s1, acquiring temperatures of a plurality of temperature control points in the computing equipment again, and determining temperature control updating errors and temperature index updating parameters according to the acquired temperatures of the plurality of temperature control points in the computing equipment again; s2, if the temperature index updating parameters meet the adjustment conditions, acquiring updating control parameters according to the temperature control updating errors and the control model; s3, controlling the operation of the temperature regulation module based on the updated control parameters; and after controlling the operation of the temperature regulation module based on the update control parameter, performing iterative processing according to S1, S2 and S3 until the temperature control update error does not meet the adjustment condition.

In this implementation manner, at a moment after the operation of the temperature adjustment module is controlled, whether to continue the adaptive adjustment of the parameter is judged again by the temperature index parameter related to the newly collected temperature (that is, the temperature index update parameter), and so on until the adaptive adjustment of the parameter is determined to be unnecessary (the adjustment condition is not satisfied) according to the temperature index parameter, so after one or more cycles, the appropriate PID parameter is adjusted to obtain the appropriate control parameter to control the operation of the temperature adjustment module, the temperature of the computing device is adjusted to be within an appropriate range or below a certain temperature threshold, and the adjustment manner of the PID parameter is more flexibly realized, which is applicable to more relevant scenes of heat dissipation and refrigeration.

In one implementation manner of this embodiment, the control model is obtained based on a PID control strategy and a neuron learning algorithm, and the determining the target control parameter according to the temperature control error and the control model includes: determining a PID parameter adjustment amount according to the recorded historical control parameter, the recorded PID parameter historical adjustment amount used for determining the historical control parameter and the temperature control historical error and the temperature control error; and obtaining control parameters through a neuron learning algorithm according to the configured PID parameters and the PID parameter adjustment quantity. Either the target control parameters or the process control parameters mentioned above may be obtained in this way.

In the implementation mode, the PID control strategy is combined with the neuron learning algorithm, the control parameter at the current moment is obtained by issuing the PID parameter adjustment quantity, the control parameter, the temperature control error at the historical moment and the temperature control error at the current moment, and the self-adaption of the PID parameter can be completed more quickly through actual measurement, so that the control of a temperature regulation module such as a fan is realized quickly, and the temperature regulation under various scenes is realized.

In one implementation of this embodiment, the determining the temperature control error according to the temperatures of the plurality of temperature control points in the computing device includes: acquiring the temperature of each temperature control point in the computing equipment, and obtaining the temperature deviation of each temperature control point according to the temperature of each temperature control point and the expected temperature; and carrying out weighted summation according to the weight value corresponding to each temperature control point and each temperature deviation to obtain a temperature control error. The expected temperature is a preset value, and the server operates at the expected temperature, so that good working performance can be maintained.

In the implementation mode, the temperature control error calculated by adopting the mode can better reflect the temperature condition of the computing equipment, and the obtained temperature control error can be better used for calculating subsequent control parameters so as to better control the temperature and realize the cooling control in the computing equipment.

In one implementation manner of this embodiment, before performing weighted summation according to the weight value corresponding to each temperature control point and each temperature deviation, before obtaining the temperature control error, the method further includes: setting weight values corresponding to the temperature control points according to the attribute information of the heating source modules corresponding to the temperature control points; wherein, the attribute information of the heating source module includes: either or both of the type of heat source module and the distance between the heat source module and the temperature regulation module in the computing device.

In the implementation mode, different calculation weights are set for different heating sources, so that temperature control error calculation can be performed on the heating sources more specifically, further control parameters for cooling control on the computing equipment are better obtained, and cooling in the computing equipment is better realized.

In one implementation manner of this embodiment, the temperature index parameter is used to represent a deviation degree of the temperature corresponding to each temperature control point from the desired temperature.

In the implementation manner, the deviation degree of the corresponding temperature of the temperature control point and the expected temperature is used as a judgment factor to judge whether the PID parameter is adaptively adjusted, so that the control model can be timely triggered to acquire the required control parameter, and the computing equipment is better cooled.

In one implementation of this embodiment, the temperature index parameter includes: and calculating the obtained variance according to the temperature corresponding to each temperature control point and the expected temperature.

In the implementation manner, whether the PID parameter is adaptively adjusted is judged by taking the variance of the corresponding temperature of the temperature control point and the expected temperature as a judgment factor, so that the control model can be timely triggered to acquire the required control parameter, and better cooling in the computing equipment is facilitated.

In one implementation manner of this embodiment, the neuron learning algorithm is a single neuron learning algorithm, an input of the single neuron learning algorithm is the configured PID parameter, the PID parameter adjustment amount is a weight value, and an output of the single neuron learning algorithm is a control parameter; the single neuron learning algorithm sums the proportion P parameter included in the configured PID parameter and the proportion P parameter adjustment amount in the PID parameter adjustment amount, sums the integral I parameter included in the configured PID parameter and the integral I parameter adjustment amount in the PID parameter adjustment amount, sums the differential D parameter included in the configured PID parameter and the differential D parameter adjustment amount in the PID parameter adjustment amount, and predicts the target control parameter through a prediction function. The control parameter output by the single neuron learning algorithm may be a target control parameter or a process control parameter, from which the target control parameter is further obtained.

In the implementation mode, the PID control strategy and the single neuron learning algorithm are further combined to deploy a calculation mode to acquire the control parameters, so that the calculation is simpler and more efficient.

In one implementation of this embodiment, the method further includes: and if the temperature index parameter does not meet the adjustment condition, generating a control parameter according to the configured PID parameter to control the temperature regulation module of the computing equipment to work.

In the implementation mode, the self-adaptive temperature control can be performed based on the PID parameters, the self-adaptive adjustment of the PID parameters can be avoided under the condition that the adjustment conditions are not met, the efficiency of the temperature adjustment is improved, and the automatic and intelligent requirements of the self-adaptive adjustment are met.

In one implementation manner of this embodiment, the determining, according to the temperature control error and the control model, a target control parameter includes: obtaining target control parameters according to the temperature control error and the neural network model; the neural network model is obtained after monitoring training based on temperature training parameters in a plurality of scenes and control parameters measured through experiments in each scene. Different scenarios correspond to different device temperature control requirements, such as a scenario including a server under different workloads (full load or half load, etc.), a scenario under different aging degrees, and so on.

In the implementation mode, the neural network model obtained based on the supervision data training by the test measurement can obtain the control parameters more quickly, efficiently and accurately, and the temperature control of the computing equipment is realized.

In one implementation of this embodiment, the method further includes: acquiring temperature change information, wherein the temperature change information comprises change values before and after controlling the operation of the temperature regulation module; if the temperature change information meets the temperature rise condition of the computing equipment, shortening the preset duration; or if the temperature change information meets the temperature reduction condition of the computing equipment, the preset duration is increased.

In the implementation mode, the preset time length for controlling the temperature regulation module to work can be dynamically regulated according to the needs, so that the self-adaptive temperature control requirement of the computing equipment can be met to a certain extent, the calculated amount can be reduced by increasing the time interval and the like, and the energy consumption is saved.

In one implementation manner of this embodiment, each temperature control point in the computing device corresponds to at least one heat source module, and the heat source module includes any one or more of a central processor, a graphics processor, and an image processing unit that are disposed in the computing device.

In the implementation mode, the computing equipment can be subjected to more comprehensive temperature monitoring, and from the more comprehensive temperature monitoring, the self-adaptive adjustment of the control parameters can be better realized, and the computing equipment can be better subjected to temperature control.

In a second aspect, an embodiment of the present application provides a control device for a temperature adjustment module, including:

an acquisition unit for acquiring a temperature of a temperature control point in the computing device;

the processing unit is used for determining temperature control errors and temperature index parameters according to the temperatures of a plurality of temperature control points in the computing equipment; if the temperature index parameter meets the adjustment condition, determining a target control parameter according to the temperature control error and a control model; and controlling the operation of the temperature regulating module based on the target control parameter.

In a third aspect, embodiments of the present application provide a computing device comprising: a processor for implementing the control method of the temperature adjustment module of the first aspect as described above.

In a fourth aspect, embodiments of the present application also provide a computing device, including: the system comprises a state detection module, a self-adaptive learning module, a PID control module and a temperature regulation module, wherein: the state detection module is used for determining temperature control errors and temperature index parameters according to the temperatures of a plurality of temperature control points in the computing equipment, and notifying the self-adaptive learning module through a trigger signal when the temperature index parameters meet adjustment conditions; the self-adaptive learning module is used for acquiring PID parameter adjustment according to the temperature control error when receiving a trigger signal; the PID control module is used for determining target control parameters according to the PID parameter adjustment quantity and the control model and controlling the temperature regulation module based on the target control parameters; the temperature adjusting module is used for executing refrigeration processing under the control of the target control parameter.

In a fifth aspect, an embodiment of the present application provides a computer-readable storage medium storing a computer program, the computer program including program instructions that, when executed, cause the control method of the temperature adjustment module of the first aspect to be implemented.

In a sixth aspect, embodiments of the present application provide a computer program product comprising a computer program or instructions which, when run on a computer, cause the computer to perform the method of controlling a temperature regulation module according to the first aspect.

Drawings

FIG. 1a is a schematic diagram of a system architecture of a computing device according to an embodiment of the present application;

FIG. 1b is a schematic diagram of a temperature control system according to an embodiment of the present application;

fig. 2 is a flow chart of a control method of a temperature adjustment module according to an embodiment of the present application;

FIG. 3 is a schematic diagram of control parameters obtained by a single neuron learning algorithm;

FIG. 4 is a flow chart of a control method of another temperature adjustment module according to an embodiment of the application;

FIG. 5 is a flow chart of a control method of a temperature adjustment module according to an embodiment of the application;

Fig. 6 is a schematic structural diagram of a control device of a temperature adjustment module according to an embodiment of the present application;

FIG. 7 is a schematic diagram of a computing device provided by the present application;

fig. 8 is a schematic structural diagram of a chip according to the present application.

Detailed Description

In order to better understand the technical scheme provided by the embodiment of the present application, first, a part of technical terms related to the embodiment of the present application are introduced.

(1) PID (Proportional, integral, differential derivative)

PID control algorithm: and a control algorithm integrating three links of pro-port, integral and Differential. And according to the input deviation value, carrying out operation according to the functional relation of proportion, integral and derivative, and controlling output by an operation result. The PID control module based on the PID control algorithm consists of a proportional unit, an integral unit and a differential unit, and controls target objects such as fans based on the setting of PID parameters (Xp, xi and Xd).

(2) Device management unit

The device management unit performs functions such as component management, asset management functions, etc., such as power on and power off control, fan speed regulation, electronic tag management, etc., within the computer system. Numerous platform management functions may be integrated onto the device management unit, and by way of example, the platform management functions may include a series of monitoring and control functions, the object of operation being system hardware. Such as by monitoring the temperature, voltage, fans, power supplies, etc. of the system and making corresponding adjustments to ensure that the system is in a healthy state. The platform management function may also include logging information and log records of various hardware for prompting the user and location of subsequent questions.

It should be noted that, different equipment manufacturers may refer to the equipment management units differently, and in the embodiment of the present application, the equipment management units may be exemplified by the baseboard management controller (Baseboard Management Controller, BMC), and it is understood that the equipment management units may also have other names.

The control method for the temperature regulation module can be applied to the computing equipment with the heat source, for example, a server comprising multiple heat sources. For example, a system architecture diagram of a computing device may be as shown in fig. 1a, the computing device being an electronic device having data processing capabilities, data transceiving capabilities, and data storage capabilities. For example, the computing device may be a server such as a rack server, or the like. Wherein the computing device 10 includes a cabinet 101, components within the cabinet 101 including, but not limited to, a device management unit 102 and a plurality of power supplies 103 (3 power supplies are illustrated in fig. 1 a), the device management unit 102 being electrically connected to the plurality of power supplies 103 via a bus 104. The device management unit 102 is mainly used for monitoring, managing, etc. the computing device 10. For example, the device management unit 102 may monitor the status (e.g., temperature, voltage, etc.) of various hardware devices in the computing device. As another example, system configuration, firmware upgrade, fault diagnosis, and the like may be performed by the device management unit 102.

The power supply 103 may be a powersupply unit (PSU).

The power supply 103 may include a processor, which may include various controllers with processing functions, such as a micro control unit (MicrocontrollerUnit, MCU), a field programmable gate array (FieldProgrammableGateArray, FPGA), a complex programmable logic device (ComplexProgrammableLogicDevice, CPLD), and the like. The MCU is a micro control unit integrated with a processing unit and a peripheral module, the processing unit is a functional unit for interpreting and executing instructions, and the processing unit in the embodiment of the present application may refer to a Central Processing Unit (CPU). Peripheral modules may include, but are not limited to: random access memory (RandomAccessMemory, RAM), read-only memory (ROM), timer counter, input/output (I/O) ports, the processing unit and the peripheral module may be connected by a bus. The MCU may also be referred to as a single chip microcomputer (SingleChipMicocomputer) or a single chip microcomputer.

Bus 104 may include, but is not limited to: a controller area network (ControllerAreaNetwork, CAN) bus, an integrated circuit bus (I2C), a system management bus (SystemManagementBus, SMBus), and the like.

The system architecture of the computing device may further include: a main board 105; various chips may also be configured on motherboard 105, such as one or more CPUs (central processing unit, central processing units), one or more GPUs (graphics processing unit, graphics processors), one or more IPUs (Image Processing Unit, image processing units), and the like.

The system architecture of the computing device may also include, but is not limited to: the temperature adjusting module is a configurable component such as basic hardware such as a fan, a hard disk, a network card, a disk array (RedundantArraysofIndependentDisks, RAID) and the like. Various controllers with processing functions such as FPGA and CPLD.

In the embodiment of the present application, temperature sensors are configured in the computing device for heat sources, such as the power supply 103, each chip on the motherboard 105, and the like, and these temperature sensors are connected to the device management unit 102, and collect the temperatures of each heat source and send the temperatures to the device management unit 102, so that the device management unit 102 performs refrigeration control on the temperatures in the computing device from the received temperatures of each heat source.

The following describes a control method for a temperature adjustment module in a computing device. The control method of the temperature adjustment module may be performed by a device management unit such as a BMC or a controller.

In heat dissipation control of a multi-heat source computing device, a temperature adjustment module, such as air-cooled, water-cooled, etc., may be heat-controlled based on a PID control algorithm such that the temperature of the computing device is within a relatively suitable temperature. The process of performing heat dissipation control based on the PID control algorithm may include: firstly, temperature information is acquired through a temperature sensor and sent to the BMC; then, after the BMC receives the temperature information, if temperature control is needed, PWM (Pulse width modulation ) signals are output to the fan through a PID control algorithm, so that the rotating speed of the fan is adjusted, heat dissipation is accelerated, and the temperature is reduced. The PID parameters used by the PID control algorithm may be parameters obtained by performing PID heat dissipation control and debugging on a specific computing device in a laboratory environment, that is, the computing device may be controlled to dissipate heat by a static PID control algorithm based on fixed parameters. Based on the heat dissipation control scheme based on the PID control algorithm, the PID control method can further dynamically adjust PID parameters, so that the PID heat dissipation control scheme can adapt to more scenes, for example, under the condition that the temperature rise of internal components is increased due to relative aging of the computing equipment, the PID parameters can be adjusted in a self-adaptive mode instead of using fixed PID parameters, and the PID control algorithm can better realize the cooling control of the computing equipment.

Fig. 1b is a schematic diagram of a temperature control system according to an embodiment of the present application. The embodiment of the application can realize the self-adaptive control of the temperature regulating module such as a fan in the computing equipment, as shown in fig. 1b, the system comprises a state detecting module, a self-adaptive learning module and a PID control module, in one possible implementation way, the PID control module comprises a unit for executing a PID control algorithm based on PID parameters and an executing unit, the executing unit can be used for directly controlling the temperature regulating module, or the executing unit can be not used, the PID control module outputs the control parameters calculated based on the PID parameters, and the temperature regulating module is indirectly controlled by an external unit.

The state detection module can be used for determining temperature control errors and temperature index parameters according to the temperatures of a plurality of temperature control points in the computing equipment, and notifying the self-adaptive learning module through a trigger signal when the temperature index parameters meet adjustment conditions; the self-adaptive learning module can be used for acquiring PID parameter adjustment according to the temperature control error when receiving a trigger signal; the PID control module can be used for determining target control parameters according to the PID parameter adjustment quantity and the control model and controlling the temperature regulation module based on the target control parameters; and a temperature regulation module in the computing device is used for executing refrigeration under the control of the target control parameter. The temperature adjusting module may specifically be a refrigeration module such as air cooling, water cooling, etc. for reducing the temperature in the computing device.

In an implementation manner of this embodiment, during the system operation of the computing device, a temperature control error e (t) may be obtained by detecting an actually measured temperature and an expected temperature in a unit time through a state detection module, and a temperature index parameter may be obtained, where the temperature index parameter is used to indicate a deviation degree of a temperature corresponding to a temperature control point from the expected temperature, specifically, the temperature index parameter may refer to a temperature variance obtained after variance calculation is performed on a temperature and an expected temperature of each currently acquired temperature control point, where the temperature variance is used to determine suitability of a current PID parameter for a fan isothermal adjustment module, for example, after a configured PID parameter (the configured PID parameter may be the current PID parameter or may also be a control parameter generated by a user manually configuring) controls the temperature adjustment module to work, the obtained temperature variance greatly exceeds a set range, and indicates that the configured PID parameter is not suitable for temperature control of a server in a current environment, and at this time, the temperature index parameter may be determined to meet an adjustment condition, so that the adaptive learning module may be started to calculate, based on the temperature control error e (t), to obtain a current PID parameter adjustment quantity P (i.e.e., a current PID parameter) and a new PID parameter P (ω (t)) and a new PID parameter I (ω) are configured to be a new PID parameter) and an isothermal parameter I (ω) are configured to be adjusted). The execution unit then obtains a control signal T (T) based on u (T) which can ultimately control the fan speed. After the temperature regulation module is controlled to work by the control parameters generated by the configured PID parameters, the obtained temperature variance is in the set range, and if the temperature index parameters are determined not to meet the adjustment conditions, the operation of the configured PID parameters adaptive server is indicated, the temperature of the computing equipment can be effectively controlled, and the configured PID parameter generation control parameters can be kept to continuously control the temperature regulation module.

That is, if the configured PID parameters can effectively control the temperature of the computing device, and can adapt to the operation of the computing device, the operation of the PID parameters is maintained; if the configured PID parameters cannot effectively control the temperature of the computing equipment and are not suitable for the operation of the computing equipment, the self-adaptive learning module can be started through the trigger signal, the PID parameters are dynamically adjusted, and finally, the proper control parameters are output to control the temperature adjusting module to work, so that the corresponding temperature variance is reduced to be within a set range, and the aim of controlling the temperature in the computing equipment within a proper temperature range is finally achieved.

It should be noted that, fig. 1b is only illustrative, and in one implementation manner of this embodiment, the functions implemented by the state detection module, the adaptive learning module, the PID control module, and the execution unit may be implemented in one controller or BMC, or the state detection module, the adaptive learning module, the PID control module, and the execution unit may be disposed in the controller or BMC.

Referring to fig. 2, fig. 2 is a flowchart of a control method of a temperature adjustment module according to an embodiment of the application, where the method may be performed by a controller or a BMC. The method includes, but is not limited to, the steps of:

S201: temperature control errors and temperature index parameters are determined according to temperatures of a plurality of temperature control points in the computing device. In the embodiment of the application, the temperature control error can be obtained based on calculation of the temperature acquired at each temperature control point at the current moment. For a computing device, there may be a plurality of heat source modules, at least one heat source module for each temperature control point in the computing device, where the heat source modules include any one or more of a central processor, a graphics processor, and an image processing unit disposed in the computing device. For example, in some servers, there are heat source modules such as one or more CPUs (central processing unit, central processing units), one or more GPUs (graphics processing unit, graphics processors), one or more IPUs (Image Processing Unit, image processing units), and the like. Each heating source module can be determined as a temperature control point according to the requirement to perform temperature acquisition operation on the heating source module. For convenience of description, the temperature control error and the temperature index parameter acquired at S201 are defined as the temperature control error at the first time and the temperature index parameter at the first time. The calculation process of the temperature control error may refer to the following description, and the temperature index parameter is a parameter for indicating the deviation degree of the temperature corresponding to each temperature control point from the expected temperature, and in one implementation may specifically be a variance calculated according to the temperature corresponding to each temperature control point and the expected temperature. The desired temperature is a preset temperature and can be considered to be a suitable temperature at which the server is desired to operate.

In one implementation of this embodiment, determining the temperature control error according to the temperatures of the plurality of temperature control points in the computing device may specifically include: acquiring the temperature of each temperature control point in the computing equipment, and obtaining the temperature deviation of each temperature control point according to the temperature of each temperature control point and the expected temperature; and carrying out weighted summation according to the weight value corresponding to each temperature control point and each temperature deviation to obtain a temperature control error. That is, based on the temperatures collected at each temperature control point, first, a temperature deviation between the collected respective temperatures and a desired temperature, which may be a uniform temperature, may be calculated. Then, the temperature control errors of the temperature control points can be counted through a temperature weighting mode.

Before the temperature control error is obtained by weighting and summing, determining a weighted value of weighting calculation, and setting the weighted value corresponding to each temperature control point according to the attribute information of the heating source module corresponding to each temperature control point; wherein, the attribute information of the heating source module includes: either or both of the type of heat source module and the distance between the heat source module and the temperature regulation module in the computing device.

As shown in table 1 below, the temperature deviations, such as e11 (t), e22 (t), emun (t), etc., of the temperatures of the plurality of groups of heat source modules acquired at time t, respectively, with respect to the desired temperature, and the weighting coefficients, such as p11, p22, pmn, etc., configured for each heat source module are schematically shown. The weighting coefficient is an experience value, for a user, the weighting levels corresponding to different types of heating source modules can be set in advance according to actual demands, the weighting coefficients of the corresponding levels can be automatically configured by different weighting levels, for example, the user can set the first weighting level for the CPU, the second weighting level for the GPU and the third weighting level for the IPU according to demands, so that the first weighting coefficient can be automatically set for the CPU according to the first weighting level, the second weighting coefficient can be automatically set for the GPU according to the second weighting level, the third weighting coefficient can be automatically set for the CPU according to the third weighting level, the first weighting coefficient, the second weighting coefficient and the third weighting coefficient are different, for example, the demands of the user are focused on the temperature of the CPU, the CPU can be set as the first weighting level, the temperature weight of the CPU is increased, and the first weighting coefficient corresponding to the CPU is a larger value. In one implementation manner of the embodiment, based on the weight level set by the user, a weighting coefficient can be automatically allocated to each level of temperature regulation module in the computing device, so that the temperature deviation of the temperature control point corresponding to each heating source module is weighted and summed to obtain the temperature control error of the computing device.

For the computing device, the weighting coefficient may be further determined based on a distance between the heat source module and the temperature adjustment module in the computing device, where a heat source closer to the temperature adjustment module sets a higher weighting coefficient and a heat source farther from the temperature adjustment module sets a lower weighting coefficient. For each heating source module in the computing device, a user can record the identification of each heating source module in a distance table or an identification sequence based on a mode that the distance between the heating source module and the temperature regulation module is from far to near or from near to far, so that a weighting coefficient can be automatically distributed for each temperature regulation module based on the distance table or the identification sequence, and the temperature deviation of a temperature control point corresponding to each heating source module can be conveniently weighted and summed to obtain the temperature control error of the computing device.

In an implementation manner of this embodiment, the weight level set for each type of heat source module and the distance table or the identification sequence set forth above may also be comprehensively considered to automatically allocate a weighting coefficient to each temperature adjustment module, so as to perform weighted summation calculation on the temperature deviation of the temperature control point corresponding to each heat source module, thereby obtaining the temperature control error of the computing device.

The weighting coefficient may be a proportional relationship, and the final weight value may be obtained through normalization processing.

TABLE 1

The temperature control error, i.e., the temperature control error e (t), can be obtained based on the following formula by the table.

S202: and if the temperature index parameter meets the adjustment condition, determining a target control parameter according to the temperature control error and the control model. The temperature index parameters include the aforementioned temperature variances with respect to the computing device, and the required temperature variances may be calculated based on the difference between the temperature of each temperature control point and the desired temperature. Whether the temperature index parameter meets the adjustment condition is mainly determined by comparing the temperature variance with a variance threshold, namely, whether the temperature variance is larger than or equal to the set variance threshold or not is determined according to the comparison result, and whether the temperature index parameter meets the adjustment condition or not is determined according to the comparison result. If the comparison result is that the temperature variance at the first moment is greater than or equal to the set variance threshold, the temperature index parameter can be considered to meet the adjustment condition, a new PID parameter adjustment amount needs to be adaptively generated to adjust the PID parameter, and the control parameter of the isothermal adjustment module capable of controlling the fan is obtained on the basis of the adjusted PID parameter. Otherwise, if the temperature variance is smaller than the set variance threshold, the temperature index parameter can be considered to not meet the adjustment condition, the PID parameter is not required to be adjusted, a new PID parameter adjustment amount is not required to be calculated, and a control parameter can be directly generated through the PID parameter corresponding to the condition that the current temperature index parameter does not meet the adjustment condition, so that the temperature adjustment module of the fan is controlled.

In one implementation of this embodiment, the control model may predict the current control parameters for controlling the fan based on the relevant parameters (which may be considered as some stored historical parameters) at a time prior to the current time, and the current temperature control error. The control model may be obtained based on a PID control strategy and a neuron learning algorithm, and the step of determining the target control parameter according to the temperature control error and the control model may include: determining a PID parameter adjustment amount according to the recorded historical control parameter, the recorded PID parameter historical adjustment amount used for determining the historical control parameter and the temperature control historical error and the temperature control error; and obtaining control parameters through a neuron learning algorithm according to the configured PID parameters and the PID parameter adjustment quantity. The control parameter finally obtained by the neuron learning algorithm may be the target control parameter or may be a process control parameter, and the target control parameter is finally obtained on the basis of the process control parameter.

Meanwhile, after each control parameter is generated, the generated control parameter, the PID parameter adjustment quantity corresponding to the generated control parameter and the temperature control error can be stored into a memory as a history record, so that when the control parameter corresponding to a new moment needs to be generated, the needed history control parameter, the corresponding PID parameter history adjustment quantity and the temperature control history error can be directly obtained from the history record to generate the PID parameter adjustment quantity, and the new control parameter can be obtained. In the history record stored in the memory, the latest acquired control parameter is taken as a history control parameter, the PID parameter adjustment quantity corresponding to the latest acquired control parameter is taken as a PID parameter history adjustment quantity, and the temperature control error corresponding to the latest acquired control parameter is taken as a temperature control history error.

The temperature control error and the temperature index parameter acquired in S201 are defined as the temperature control error and the temperature index parameter of the first time, and then S202 may specifically include: obtaining the PID parameter adjustment amount at the first time based on the PID parameter adjustment amount at the third time (PID parameter history adjustment amount), the control parameter predicted at the third time (history control parameter), the temperature control error at the first time and the temperature control error at the third time (temperature control history error); and obtaining the control parameters at the first moment through a neuron learning algorithm according to the configured PID parameters and the PID parameter adjustment quantity at the first moment. The third time is prior to the first time. The PID parameter adjustment amount at the first moment is calculated as follows.

Taking the first moment as the t moment as an example, at the third moment (t-1 moment) which is the moment before the first moment, the corresponding PID parameter adjustment quantity and control parameters are calculated, the PID parameter adjustment quantity at the t moment is calculated based on the data at the t-1 moment, and further the control parameters at the t moment are obtained based on the PID parameter adjustment quantity at the t moment and the existing PID parameters (namely the configured PID parameters).

In an implementation manner of this embodiment, the neuron learning algorithm may be a single neuron learning algorithm, where an input of the single neuron learning algorithm is a configured PID parameter, a PID parameter adjustment amount at a first time is a weight value, and an output of the single neuron learning algorithm is a control parameter at the first time, where the configured PID parameter is a default PID parameter or a PID parameter calculated at a third time, where the PID parameter at the third time may be obtained by summing the PID parameter calculated at the last time at the third time and the PID parameter adjustment amount calculated at the third time, and so on; the single neuron learning algorithm sums the proportion P parameter included in the configured PID parameter and the proportion P parameter adjustment amount in the PID parameter adjustment amount at the first moment, sums the integral I parameter included in the configured PID parameter and the integral I parameter adjustment amount in the PID parameter adjustment amount at the first moment, sums the differential D parameter included in the configured PID parameter and the differential D parameter adjustment amount in the PID parameter adjustment amount at the first moment, and predicts the control parameter at the first moment through a prediction function.

The above-mentioned PID parameter adjustment can be calculated by the following formula:

ωp (t) =ωp (t-1) +ηp×e (t) ×u (t-1) ×1 (t) formula 2;

ωi (t) =ωi (t-1) +ηi×e (t) ×u (t-1) ×2 (t) formula 3;

ωd (t) =ωd (t-1) +ηd×e (t) ×u (t-1) ×3 (t) formula 4;

wherein, x1 (t) = (t) - (t-1), x2 (t) = (t), x3 (t) = (t) -2e (t-1) +e (-2).

ηp, ηi and ηd are the learning speeds of the proportional, integral and differential terms, respectively, and represent a learning factor, which can be defined according to empirical values. In one implementation of this embodiment, a small value, such as a value of 0.1 or less, may be set, the smaller the setting, the slower the learning rate, but not easily diverged. If a fast learning rate is not required, a smaller value may be set but not too small;

e (t) is the temperature control error mentioned above, and represents the difference between the measured temperature at time t and the expected temperature, and can be obtained by calculation according to formula 1, e (t-1) represents the difference between the measured temperature at time t-1 and the expected temperature, and e (t-2) represents the difference between the measured temperature at time t-2 and the expected temperature;

u (t-1) is a control parameter at time t-1. The control parameter is an adjustment amount for controlling the rotation speed of the fan, the fan system is generally controlled by PWM, and the rotation speed of the fan is determined by the duty cycle, for example, 40% of the duty cycle may correspond to 2000 rotation speed, and 60% of the duty cycle may correspond to 3000 rotation speed. u (t) is a parameter related to the duty cycle. It will be appreciated that this does not represent u (t) =60%, u (t) being just a value such as 126 etc. and being subsequently functionally converted to the corresponding duty cycle.

Omega P (t) is the proportional P parameter adjustment amount at time t, omega P (t-1) is the proportional P parameter adjustment amount at time t-1, omega I (t) is the integral I parameter adjustment amount at time t-1, omega I (y-1) is the integral I parameter adjustment amount at time t-1, omega D (t) is the D parameter adjustment amount at time t, and omega D (t-1) is the D parameter adjustment amount at time t-1.

In one implementation manner of this embodiment, firstly, a PID parameter adjustment amount is calculated based on a temperature control error e (t) at a first moment, then, based on an existing PID parameter, a control parameter is obtained through a single neuron learning algorithm by taking the PID parameter adjustment amount as a weight value, and specifically referring to fig. 3, fig. 3 is a schematic diagram of the control parameter obtained through the single neuron learning algorithm, and the control parameter u (t) is obtained by calculating the configured PID parameter through the single neuron learning algorithm by taking the PID parameter adjustment amount as a weight value. As can be seen from fig. 3, the inputs of the single neuron learning algorithm are the configured PID parameters (Xp, xi, and Xd in fig. 3), and the PID parameter adjustment amounts calculated by the above-described formulas 2, 3, and 4 are calculated as weight values. The configured PID parameters may be default PID parameters (default proportional P parameter, integral I parameter, D parameter), which may be initial PID parameters that are initially set, and each time after obtaining the PID parameter adjustment, the PID parameter obtained by adding the PID parameter adjustment to the default PID parameter is used as a new default PID parameter, and a default PID parameter is covered.

For the embodiment of the application, if the temperature index parameter meets the adjustment condition, the PID parameter adjustment quantity is obtained through a formula 1, a formula 2, a formula 3 and a formula 4, and the PID parameter adjustment quantity is used as a weight value on the basis of the configured PID parameter, and the control parameter is obtained through a single neuron learning algorithm; if the temperature index parameter does not meet the adjustment condition, the PID parameter adjustment quantity obtained by calculation through the formula is not needed, and the control parameter is directly generated based on the PID parameter at the last moment and the PID parameter obtained by PID parameter adjustment at the last moment.

Or if the temperature index parameter meets the adjustment condition, obtaining PID parameter adjustment quantity through a formula 1, a formula 2, a formula 3 and a formula 4, and obtaining control parameters through a single neuron learning algorithm by taking the PID parameter adjustment quantity as a weight value on the basis of the configured PID parameter; if the temperature index parameter does not meet the adjustment condition, the PID parameter adjustment quantity calculated by the formula is not needed, the PID parameter at the previous moment is still used as input, the PID parameter adjustment quantity obtained at the previous moment is used as a weight value, the control parameter is obtained by a single neuron learning algorithm, and the method is equivalent to directly generating the control parameter by using the original PID parameter.

In an implementation manner of this embodiment, the determining, according to the temperature control error and the control model, a target control parameter may also include: obtaining target control parameters according to the temperature control error at the first moment and the neural network model; the neural network model is obtained after monitoring training based on temperature training parameters in a plurality of scenes and control parameters measured through experiments in each scene. That is, the control model may directly obtain a control parameter based on the temperature control error at the first time, for example, the control model corresponds to a neural network model obtained by training through a large number of temperature control errors in a scene and a supervision tag measured through experiments in a laboratory or the like, and after training is completed, the neural network model may take data of the temperature control errors, the temperature and the like as input and output the data as the control parameter corresponding to the time.

S203: and controlling the temperature regulating module to work based on the target control parameter. After the control parameter u (t) is obtained, the temperature adjustment module is controlled with the control parameter u (t) as an adjustment amount of the temperature control to desirably reduce the temperature of the computing device, for example, the rotation speed of the fan is adjusted based on the adjustment amount of the temperature control, and the rotation speed of the fan is increased so as to reduce the temperature of the computing device.

In one implementation manner of this embodiment, after S201 to S203 are executed, the control method of the temperature adjustment module of the present application may be implemented, and the operation of the temperature adjustment module may be controlled by adaptively generating relatively suitable control parameters based on the control model, starting from the temperatures of the respective temperature control points in the computing device dynamically. In other embodiments, the following step one and subsequent steps may be further performed to further improve the adaptability of the PID parameter adjustment.

In an implementation manner of this embodiment, the step S202 of determining the target control parameter may be a continuous optimization adjustment process, and the step S202 may include: if the temperature index parameter meets the adjustment condition, iterating the process control parameter at least once according to the temperature control error and the control model; wherein the process control parameter is configured to control the temperature adjustment module to operate for a preset duration; and if the temperature regulation module works based on the latest process control parameter so that the temperature index parameter does not meet the regulation condition, determining the latest process control parameter as the target control parameter. Before determining the final target control parameter, iterating according to the steps of S201-S203 to obtain the process control parameter, and when the obtained temperature index parameter does not meet the adjustment condition after a certain process control parameter controls the temperature adjustment module to work, the process control parameter at the moment is the target control parameter.

Taking the first time (t) and the second time (t+1) as an example, after executing the above S201, S202 and S203 at the time t, if the temperature index parameter obtained again still satisfies the adjustment condition, the control parameter obtained at the time t is a process control parameter, and the above S201, S202 and S203 are continuously executed at the time t+1 until the temperature index parameter does not satisfy the adjustment condition after newly executing the above S201, S202 and S203, at this time, the control parameter obtained at the time t+1 is determined as the target control parameter, and the cooling control is performed on the temperature adjustment module.

In another implementation manner of this embodiment, after step S203 is performed, the method may further include: step one, acquiring temperatures of a plurality of temperature control points in the computing equipment again, and determining temperature control updating errors and temperature index updating parameters according to the acquired temperatures of the plurality of temperature control points in the computing equipment again. For ease of understanding, the time at which the temperatures of the plurality of temperature control points within the computing device are again acquired may be understood as a second time, which is subsequent to the first time. That is, after S203 is performed, for example, after the rotational speed of the fan is increased to cool down, the temperatures acquired by the respective temperature control points are acquired again, and the temperature control error at the second time (i.e., the temperature control update error in step one) and the temperature index parameter at the second time (temperature index update parameter) are acquired in the same manner as described above, so as to determine whether to continue the process operation of generating new control parameters to adjust the temperature adjustment module.

Step two, if the temperature index updating parameters meet the adjustment conditions, updating control parameters are obtained according to the temperature control updating errors and the control model. That is, if the temperature index parameter at the second time meets the adjustment condition, a new control parameter is obtained again according to the temperature control error and the control model at the second time, so as to obtain the control parameter (updated control parameter) at the second time, so that the temperature adjustment module of the equipment can be controlled to work through the control parameter at the second time.

And step three, controlling the operation of the temperature regulation module based on the updated control parameters. And after controlling the operation of the temperature regulation module based on the update control parameter, performing iterative processing according to the first step, the second step and the third step until the temperature control update error does not meet the adjustment condition.

In one implementation manner of this embodiment, the process of obtaining the updated control parameter according to the temperature control update error and the control model may refer to the foregoing description, and may be specifically understood that the second time is taken as the new first time, and the relevant steps of determining the PID parameter adjustment amount and finally obtaining the control parameter are repeatedly performed. Meanwhile, after the execution of the third step, the first step is continued at a time after the second time so as to determine whether to continue the subsequent steps, etc. The above steps such as the calculation and processing of the related parameters by the formulas 2, 3 and 4 are not needed to be executed again until the temperature index parameter obtained at a certain time does not satisfy the adjustment condition. In an implementation manner of this embodiment, when the temperature index parameter or the temperature index update parameter does not meet the adjustment condition, the fourth control parameter may be directly generated according to the configured PID parameter to control the temperature adjustment module of the computing device to work.

In one implementation manner of this embodiment, after controlling the operation of the temperature adjustment module based on the target control parameter, the executing of the temperature control according to the plurality of temperature control points in the acquired computing device is triggered based on the set time interval parameter, and a temperature control update error and a temperature index update parameter are determined; the method may further comprise: acquiring temperature change information, wherein the temperature change information comprises change values before and after controlling the operation of the temperature regulation module; if the temperature change information meets the temperature rise condition of the computing equipment, shortening the time interval indicated by the time interval parameter; or if the temperature change information meets the temperature reduction condition of the computing equipment, increasing the time interval indicated by the time interval parameter. That is, by simply calculating the temperature change, it is quickly determined whether a new round of PID parameter adjustment is required as soon as possible, i.e., whether the above steps one to two are performed again as soon as possible. The temperature of the computing device can be acquired by a single temperature acquisition module and the temperature change information can be determined, or the temperature change information can be determined according to the temperature acquired in real time or in a period by a module for acquiring the temperatures of a plurality of temperature control points in the computing device, once the temperature rise or fall of the computing device is determined according to the temperature change information, the time interval parameter can be modified, and the time point for starting to execute the first step is determined according to the modified time interval parameter.

Taking the first time and the second time as examples, where the second time is determined according to the first time and the time interval parameter, the parameter values of the time interval parameter between different times are allowed to be dynamically set, when the temperature change information meets the temperature rising condition, the time interval parameter is reduced, for example, the temperature change information may include a temperature difference, a temperature variance, a temperature control error, and the like, and if the temperature difference obtained by subtracting the temperature before the control from the temperature after the operation of the temperature adjustment module is controlled is greater than a preset temperature threshold, or the difference between the actually measured temperature and the expected temperature is greater than the preset temperature threshold, or the temperature variance is greater than the preset variance threshold, or the temperature control error is greater than the preset error threshold, the set time interval is reduced, and otherwise, the set time interval is increased.

It can be understood that the main scheme of the application is as follows: whether self-adaptive adjustment of controller parameters such as PID is carried out or not is judged through the collected temperature-related temperature index parameters, namely whether self-adaptive adjustment is needed or not is judged, namely whether the PID parameter adjustment quantity is needed to be updated or not to adjust the PID parameters so as to obtain new control parameters for realizing temperature adjustment, if self-adaptive adjustment is needed, the PID parameter adjustment quantity is determined based on the current collected temperature-related temperature control errors so as to obtain new control parameters, and the temperature adjustment module such as a fan is controlled to work through the new control parameters. The control parameters of the control model can be predicted by a PID control strategy and a neuron learning algorithm, and the control parameters can be predicted by directly taking some trained neural network models as the control model, for example, the inputs of the neural network models comprise the temperature control errors (of course, the inputs can also comprise the PID parameters configured in the prior art, the directly acquired temperature and the like, and the specific inputs are related to the training of the neural network models) and are output as the control parameters for controlling the temperature regulation module. In the training process of the neural network model, temperature control errors and/or other data under a plurality of scenes (such as a plurality of scenes with different numbers of temperature control points, a plurality of scenes with different fan aging degrees and a plurality of scenes with different types of temperature regulation modules) can be obtained as training data, and corresponding supervision labels (one manually regulated and suitable for the control supervision parameters of the corresponding scenes) under each scene are set in advance in a manual (laboratory manual regulation) mode and the like so as to train the initial neural network model, so that the neural network model which can be used for predicting the control parameters of the first moment through the temperature control errors of the first moment is optimized.

In the embodiment of the application, on one hand, the temperature control of the multi-heat-source computing equipment can be realized through a PID control strategy, and on the other hand, the conditions of aging of a temperature adjusting module such as a fan, environmental temperature change, equipment configuration change, operation business change and the like can occur in the use process of the computing equipment, and the changes possibly cause the capability range of the temperature control scheme preset in the computing equipment to be subjected to effective temperature control to be exceeded, so that the self-adaptive adjustment of PID parameters and the generation of control parameters can be realized on the basis of a static PID control strategy.

Further, the computing device may be in a situation beyond the laboratory simulation during use, which may result in an ineffective control of temperature, giving a temperature alarm, and may result in performance degradation of the computing device or even damage to components. If the service personnel manually adjusts parameters according to different use scenes to adapt to a new environment, the problem of low efficiency exists, and in some complex situations, the parameter adjustment cannot be completed even, so that the aim of effectively controlling the temperature of the computing equipment cannot be achieved. The application can dynamically adjust the PID parameter adjustment amount and PID parameter according to the temperature of the actual environment, realizes the self-adaption of the controller parameter to effectively cope with the corresponding scene change, and calculates the control parameter by repeatedly probing and predicting and adjusting to obtain the proper PID parameter adjustment amount and PID parameter under the current scene, thereby controlling the work of the isothermal adjustment module of the fan. The adaptability of heat dissipation control to component change, environment change and heating source change is improved, a stable and reliable temperature environment is provided for the computing equipment, and the efficiency of temperature control adapting to different scenes is improved.

The control method of the present application will be described with reference to fig. 1b and 4. Fig. 4 is a flowchart of another control method of a temperature adjustment module according to an embodiment of the application. The method of the embodiment of the present application may be implemented by a BMC or a controller, and in one implementation manner of this embodiment, each module such as that shown in fig. 1b may be set in the BMC or the controller to implement a control method of the temperature adjustment module. The embodiment of the application mainly aims at the situation that the original parameter curing temperature control scheme can not be well adapted to the scenes such as the aging of the temperature adjusting module such as a fan, the change of parts due to damage, environmental change and the like. The method of the embodiment of the application comprises the following steps.

S401: the temperature control error and the temperature index parameter at the first moment are obtained, and in the embodiment of the application, the temperature control error is obtained by calculating the temperature acquired based on a plurality of temperature control points, and can be realized by a state detection module in fig. 1 b. Meanwhile, whether the subsequent parameter adaptive processing is performed may be determined by calculating the obtained temperature index parameter, in an implementation manner of this embodiment, a variance value range may be set, where the temperature index parameter is a temperature variance calculated based on the temperature of each temperature control point, and if the temperature variance exceeds a variance threshold (a maximum value of the variance value range) of the variance value range, the temperature index parameter at the first moment may be considered to satisfy an adjustment condition, and S402 is executed. If the temperature variance is within the variance value range, S404 is directly performed, and if the temperature variance is smaller than the minimum value of the variance range, no processing may be performed, and after a certain time interval (for example, after the time interval indicated by the time interval parameter mentioned in the above embodiment), S401 may be performed again. S401 corresponds to S201 in the foregoing embodiment.

S402: starting an adaptive learning module; the self-adaptive learning module can finally calculate and obtain the PID parameter adjustment quantity through the formula 1, the formula 2, the formula 3 and the formula 4.

S403: and obtaining the PID parameter adjustment quantity calculated by the self-adaptive learning module. S403 to S405 may be performed at a PID control module as shown in fig. 1 b.

S404: and combining the configured PID parameters with the PID parameter adjustment quantity to obtain new PID parameters. It is understood that, for S404, if it is determined that S404 can be directly performed according to the temperature index parameter obtained in S401, then in this case, the configured PID parameter may be used continuously; if S402 and S403 are first performed according to S401, the PID parameter adjustment amount is the PID parameter adjustment amount obtained in S403.

S405: and calculating a target control parameter, namely a control parameter u (t) at a first moment, for the obtained new PID parameter, and outputting u (t).

It will be appreciated that S402 to S405 described above are associated with S202 in the foregoing embodiment. Specifically, specific implementations of S402 and S403 may refer to the acquisition process with respect to ωp (t), ωi (t), ωd (t) in the foregoing embodiment, and implementations of S404 and S405 refer to the process of obtaining the control parameter by the single neuron learning algorithm.

S406: the rotational speed of the fan is controlled by the execution unit based on the target control parameter. S405 and S406 may be implemented in a PID control module.

When the next time after the execution of S406, that is, when the second time comes, the second time is taken as a new first time, the above steps related to S401-S406 are repeatedly executed, that is, the temperature monitoring and sampling are performed at the second time, so as to obtain a new temperature of each temperature control point, and further, the new processing is executed through the state detection module, the adaptive learning module and the PID control module, and the iterative learning is performed until the finally collected temperature index parameter does not meet the adjustment condition, so that the new PID parameter can meet a new scene, and the temperature of the computing device can be effectively controlled under the new scene, so that the temperature in the computing device is within a reasonable temperature range.

The scheme of the application can dynamically adjust PID parameters to generate control parameters so as to adapt to corresponding scene changes, and finds out the most suitable control parameters to control the isothermal adjusting module of the fan through repeated trial and error prediction under the current scene. The adaptability of the server heat radiation system to component change, environment change and heating source change is enhanced, a stable and reliable temperature environment is provided for the server, the efficiency of temperature control adapting to different scenes is improved, the manpower resource cost is reduced, and the satisfaction degree of users when using corresponding computing equipment is improved.

The control method of the present application will be described with reference to fig. 1b and 5. Fig. 5 is a flowchart of another control method of a temperature adjustment module according to an embodiment of the application. The method of the embodiment of the present application may be implemented by a BMC or a controller, and in one implementation manner of this embodiment, each module such as that shown in fig. 1b may be set in the BMC or the controller to implement a control method of the temperature adjustment module. The embodiment of the application mainly aims at the condition that the configuration operation of the new computing equipment such as a server is completed, and the PID parameter is adaptively configured from an initially set PID parameter so as to cope with the temperature control in the future possibly occurring scene such as the aging of the fan mentioned in the embodiment. The method of the embodiment of the application comprises the following steps.

S501: and acquiring a temperature control error and a temperature index parameter at the first moment.

S502: starting an adaptive learning module; the adaptive learning module can calculate the PID parameter adjustment through the above formula 1, formula 2, formula 3 and formula 4.

S503: and obtaining the PID parameter adjustment quantity calculated by the self-adaptive learning module. S503 to S505 may be performed at a PID control module as shown in fig. 1 b.

S504: and obtaining new PID parameters by combining the configured PID parameters with the PID parameter adjustment quantity.

S505: and calculating the obtained new PID parameters to obtain target control parameters. I.e. the output u (t).

S506: the rotational speed of the fan is controlled by the execution unit based on the target control parameter.

When the next time after S506 is executed, that is, when the second time arrives, the second time is taken as a new first time, and the steps related to S501 to S506 are repeatedly executed.

It will be appreciated that S501 to S506 are the same as S401 to S406 of the previous embodiment, except that the following steps are also included in the embodiment of the present application.

S507: whether the temperature index parameter at the new moment meets the adjustment condition. The temperature index parameter at the new time may refer to a temperature index parameter obtained at the next time after the execution of S501 to S506 once. If the determination at S507 is negative, the following S508 is executed, and if the determination is positive, the steps S502 and the subsequent related steps are continued.

S508: new PID parameters are obtained.

S509: and determining the acquired new PID parameters as configured PID parameters so that the configured PID parameters can be directly acquired when the temperature index parameters are determined to meet the adjustment conditions next time.

S510: and configuring the configured PID parameters in the PID control module, and then executing the step of combining the configured PID parameters with the PID parameter adjustment according to the configured PID parameters.

The existing static parameter adjustment method needs to perform thermal test on each configuration in a laboratory, manually diagnose according to test results, adjust PID parameters and retest. And the method is circulated in this way, and the adaptive parameters can be obtained after the PID parameters are regulated for a plurality of times. In the process, more manpower and material resources are consumed, and parameters are different due to the difference of personnel capacity. By adopting the embodiment of the application, only one initial value is required to be configured, the PID parameters are adaptively adjusted through automatic testing, the manpower input is reduced, and the configuration efficiency is improved.

Corresponding to the method presented in the above method embodiment, the embodiment of the present application further provides a corresponding apparatus, including a module or unit for executing the corresponding module or unit of the above embodiment. The modules or units may be software, hardware, or a combination of software and hardware.

Referring to fig. 6, a schematic structural diagram of a control device of a temperature adjustment module according to an embodiment of the present application may be applied to a controller or EMC, where the device includes:

An acquisition unit 601 for acquiring a temperature of a temperature control point in a computing device;

a processing unit 602, configured to determine a temperature control error and a temperature index parameter according to temperatures of a plurality of temperature control points in the computing device; if the temperature index parameter meets the adjustment condition, determining a target control parameter according to the temperature control error and a control model; and controlling the operation of the temperature regulating module based on the target control parameter.

In one implementation manner of this embodiment, the processing unit 602 is further configured to obtain temperatures of a plurality of temperature control points in the computing device after being configured to control the operation of the temperature adjustment module based on the target control parameter, and determine a temperature control update error and a temperature index update parameter according to the re-obtained temperatures of the plurality of temperature control points in the computing device; if the temperature index updating parameters meet the adjustment conditions, acquiring updating control parameters according to the temperature control updating errors and the control model; and controlling the operation of the temperature regulation module based on the updated control parameter.

In one implementation manner of this embodiment, the processing unit 602 is configured to, when determining the target control parameter according to the temperature control error and the control model if the temperature index parameter meets the adjustment condition, iterate the process control parameter at least once according to the temperature control error and the control model if the temperature index parameter meets the adjustment condition; wherein the process control parameter is configured to control the temperature adjustment module to operate for a preset duration; and if the temperature regulation module works based on the latest process control parameter so that the temperature index parameter does not meet the regulation condition, determining the latest process control parameter as the target control parameter.

In one implementation manner of this embodiment, the control model is obtained based on a PID control strategy and a neuron learning algorithm, and the processing unit 602 is configured to determine, when determining a target control parameter according to the temperature control error and the control model, a historical adjustment amount of the PID parameter and a temperature control historical error according to the recorded historical control parameter and the recorded PID parameter used for determining the historical control parameter, and determine a PID parameter adjustment amount according to the temperature control error; and obtaining control parameters through a neuron learning algorithm according to the configured PID parameters and the PID parameter adjustment quantity.

In one implementation manner of this embodiment, the processing unit 602 is configured to, when determining a temperature control error according to temperatures of a plurality of temperature control points in the computing device, obtain temperatures of the temperature control points in the computing device, and obtain temperature deviations of the temperature control points according to the temperatures of the temperature control points and an expected temperature; and carrying out weighted summation according to the weight value corresponding to each temperature control point and each temperature deviation to obtain a temperature control error.

In one implementation manner of this embodiment, before the processing unit 602 is configured to perform weighted summation according to the weight values corresponding to the temperature control points and the temperature deviations to obtain the temperature control errors, the processing unit is further configured to set the weight values corresponding to the temperature control points according to the attribute information of the heat source modules corresponding to the temperature control points; wherein, the attribute information of the heating source module includes: either or both of the type of heat source module and the distance between the heat source module and the temperature regulation module in the computing device.

In one implementation manner of this embodiment, the neuron learning algorithm is a single neuron learning algorithm, an input of the single neuron learning algorithm is the configured PID parameter, the PID parameter adjustment amount is a weight value, and an output of the single neuron learning algorithm is a control parameter; the single neuron learning algorithm sums the proportion P parameter included in the configured PID parameter and the proportion P parameter adjustment amount in the PID parameter adjustment amount, sums the integral I parameter included in the configured PID parameter and the integral I parameter adjustment amount in the PID parameter adjustment amount, sums the differential D parameter included in the configured PID parameter and the differential D parameter adjustment amount in the PID parameter adjustment amount, and predicts the target control parameter through a prediction function.

In an implementation manner of this embodiment, the processing unit 602 is further configured to generate, if the temperature index parameter does not meet the adjustment condition, a fourth control parameter according to the configured PID parameter to control the operation of the temperature adjustment module of the computing device.

In one implementation manner of this embodiment, the processing unit 602 is configured to obtain, when determining the target control parameter according to the temperature control error and the control model, the target control parameter according to the temperature control error at the first time and the neural network model; the neural network model is obtained after monitoring training based on temperature training parameters in a plurality of scenes and control parameters measured through experiments in each scene.

In one implementation manner of this embodiment, after controlling the operation of the temperature adjustment module based on the target control parameter, the processing unit 602 triggers execution of the temperature control according to the plurality of temperature control points in the acquired computing device based on the set time interval parameter, and determines a temperature control update error and a temperature index update parameter; the processing unit 602 is further configured to obtain temperature change information, where the temperature change information includes a change value before and after controlling the operation of the temperature adjustment module; if the temperature change information meets the temperature rise condition of the computing equipment, shortening the time interval indicated by the time interval parameter; or if the temperature change information meets the temperature reduction condition of the computing equipment, increasing the time interval indicated by the time interval parameter.

Based on the same inventive concept, the principle and beneficial effects of the control device for a temperature adjustment module provided in the embodiments of the present application are similar to those of the embodiments of the methods of the present application, and may be referred to the principle and beneficial effects of the embodiments of the methods, and are not described herein for brevity.

Referring to fig. 7, a schematic structural diagram of a computing device according to the present application includes at least: and a processor for implementing the control method for the temperature adjustment module in the above embodiments. In one implementation of the present embodiment, the computing device 70 shown in fig. 7 includes at least one processor 701 and a transceiver 702, and optionally, may also include a memory 703.

The memory 703 may be a volatile memory, such as a random access memory; the memory may also be a non-volatile memory such as, but not limited to, read-only memory, flash memory, hard disk (HDD) or Solid State Drive (SSD), or the memory 703 is any other medium that can be used to carry or store desired program code in the form of instructions or data structures and that can be accessed by a computer. The memory 703 may be a combination of the above.

The specific connection medium between the processor 701, the transceiver 702, and the memory 703 is not limited in the embodiments of the present application. In the embodiment of the present application, the processor 701, the transceiver 702 and the memory 703 are connected by the bus 704 in fig. 7, and the bus 704 is shown by a thick line in fig. 7, and the connection manner between other components is only schematically illustrated, but not limited thereto. The bus 704 may be divided into an address bus, a data bus, a control bus, and the like. For ease of illustration, only one thick line is shown in fig. 7, but not only one bus or one type of bus.

The processor 701 may have a data transceiving function, and may be capable of communicating with other devices or apparatuses (for example, communicating between the CPLD and the BMC through localbus), and in fig. 7, a separate data communication unit may also be provided, for example, a transceiver 702, for transceiving data; the processor 701 may communicate with other devices or apparatuses via the transceiver 702 for data transmission.

In one example, when the computing device is in the form shown in FIG. 7, the processor in FIG. 7 may perform any of the method embodiments described above.

Specifically, the functions/implementation procedures of the acquiring unit and the processing unit of fig. 6 may be implemented by the processor 701 in fig. 7 calling the computer executing instructions stored in the memory 703. Alternatively, the functions/implementation of the acquisition unit of fig. 6 may be implemented by the transceiver 702 in fig. 7, and the functions/implementation of the processing unit may be implemented by the processor 701 in fig. 7 calling computer-executable instructions stored in the memory 703.

In one implementation, computing device 70 may also include circuitry that may implement the relevant functions of the method embodiments described above. The processor described in the present application may be implemented as follows: integrated circuits (integrated circuit, IC), analog ICs, radio frequency integrated circuits RFIC, mixed signal ICs, application specific integrated circuits (application specific integrated circuit, ASIC), printed circuit boards (printed circuit board, PCB), computing devices, and the like. The processor may also be fabricated using the following IC process techniques: such as complementary metal oxide semiconductor (complementary metal oxide semiconductor, CMOS), N-type metal oxide semiconductor (NMOS), P-type metal oxide semiconductor (positive channel metal oxide semiconductor, PMOS), bipolar junction transistor (bipolar junction transistor, BJT), bipolar CMOS (BiCMOS), silicon germanium (SiGe), gallium arsenide (GaAs), etc.

The computing device described in the above embodiments may include a Controller (CPLD) or a device management unit (BMC), and the structure of the computing device may not be limited by fig. 7. The computing device may be a stand-alone device or may be part of a larger device. For example, the computing device may be:

(1) A stand-alone integrated circuit IC, or chip, or a system-on-a-chip or subsystem;

(2) A set of one or more ICs, optionally including storage means for storing data, a computer program;

(3) An ASIC, such as a Modem (Modem);

(4) Modules that may be embedded within other devices;

(5) Receivers, terminals, smart terminals, cellular telephones, wireless devices, handsets, mobile units, vehicle devices, network devices, cloud devices, artificial intelligence devices, etc.;

(6) Others, and so on.

For the case where the computing device may be a chip or a system of chips, reference may be made to the schematic structural diagram of the chip shown in fig. 8. The chip shown in fig. 8 includes a processor 801 and an interface 802, and the number of the processor 801 may be one or more. The processor 801 is adapted to perform the relevant steps of the method embodiments described above.

Optionally, the chip may further comprise a memory 803, the memory 803 being for storing the necessary computer programs and data. The memory 803 may be provided separately or integrated with the processor 801.

It can be understood that some optional features of the embodiments of the present application may be implemented independently in some scenarios, independent of other features, such as the scheme on which they are currently based, so as to solve corresponding technical problems, achieve corresponding effects, or may be combined with other features according to requirements in some scenarios. Accordingly, the device provided in the embodiment of the present application may also implement these features or functions accordingly, which will not be described herein.

Those of skill in the art will further appreciate that the various illustrative logical blocks (illustrative logical block) and steps (step) described in connection with the embodiments of the present application may be implemented by electronic hardware, computer software, or combinations of both. Whether such functionality is implemented as hardware or software depends upon the particular application and design requirements of the overall system. Those skilled in the art may implement the described functionality in varying ways for the respective application, but such implementation should not be understood to be beyond the scope of the embodiments of the present application.

The described aspects of the application may be implemented in various ways. For example, these techniques may be implemented in hardware, software, or a combination of hardware. For a hardware implementation, the processing units for performing these techniques at the control device of the temperature regulation module may be implemented in one or more general purpose processors, digital signal processors (digital signal processor, DSPs), digital signal processing devices, application specific integrated circuits (application specific integrated circuit, ASICs), programmable logic devices, field programmable gate arrays (field programmable gate array, FPGAs), or other programmable logic devices, discrete gate or transistor logic, discrete hardware components, or any combinations thereof. A general purpose processor may be a microprocessor, but in the alternative, the general purpose processor may be any conventional processor, controller, microcontroller, or state machine. A processor may also be implemented as a combination of computing devices, e.g., a digital signal processor and a microprocessor, a plurality of microprocessors, one or more microprocessors in conjunction with a digital signal processor core, or any other similar configuration.

The application also provides a computer readable medium having stored thereon a computer program which when executed by a computer performs the functions of any of the method embodiments described above.

The application also provides a computer program product which, when executed by a computer, implements the functions of any of the method embodiments described above.

In the above embodiments, it may be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented in software, may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer instructions are loaded and executed on a computer, the processes or functions described in accordance with embodiments of the present application are produced in whole or in part. The computer may be a general purpose computer, a special purpose computer, a computer network, or other programmable apparatus. The computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another computer-readable storage medium, for example, the computer instructions may be transmitted from one website, computer, server, or data center to another website, computer, server, or data center by a wired (e.g., coaxial cable, fiber optic, digital subscriber line (digital subscriber line, DSL)) or wireless (e.g., infrared, wireless, microwave, etc.). The computer readable storage medium may be any available medium that can be accessed by a computer or a data storage device such as a server, data center, etc. that contains an integration of one or more available media. The usable medium may be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., high-density digital video disc (digital video disc, DVD)), or a semiconductor medium (e.g., solid state disk (solid state drive, SSD)), etc.

It is appreciated that reference throughout this specification to "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment of the present application. Thus, various embodiments are not necessarily referring to the same embodiments throughout the specification. Furthermore, the particular features, structures, or characteristics may be combined in any suitable manner in one or more embodiments. It should be understood that, in various embodiments of the present application, the sequence numbers of the foregoing processes do not mean the order of execution, and the order of execution of the processes should be determined by the functions and internal logic thereof, and should not constitute any limitation on the implementation process of the embodiments of the present application.

It will be appreciated that in the present application, the terms "when …", "if" and "if" refer to the corresponding processes that the device will perform under some objective condition, are not intended to limit the time and do not require that the device must perform a judgment, nor are other limitations intended to the scope of the application.

Elements referred to in the singular are intended to be used in the present disclosure as "one or more" rather than "one and only one" unless specifically stated otherwise. In the present application, "at least one" is intended to mean "one or more" and "a plurality" is intended to mean "two or more" unless specifically indicated.

In addition, the term "and/or" herein is merely an association relationship describing an association object, and means that three relationships may exist, for example, a and/or B may mean: there are three cases where a alone exists, where a may be singular or plural, and where B may be singular or plural, both a and B exist alone.

Those skilled in the art will understand that, for convenience and brevity, the specific working process of the system, apparatus and unit described above may refer to the corresponding process in the foregoing method embodiment, which is not described herein again.

The same or similar parts may be referred to each other in the various embodiments of the application. In the embodiments of the present application, and the respective implementation/implementation methods in the embodiments, if there is no specific description and logic conflict, terms and/or descriptions between different embodiments, and between the respective implementation/implementation methods in the embodiments, may be consistent and may refer to each other, and technical features in the different embodiments, and the respective implementation/implementation methods in the embodiments, may be combined to form a new embodiment, implementation, or implementation method according to their inherent logic relationship. The embodiments of the present application described above do not limit the scope of the present application.

The foregoing is merely illustrative of the present application, and the present application is not limited thereto, and any person skilled in the art will readily recognize that variations or substitutions are within the scope of the present application.

Claims

1. A method of controlling a temperature regulation module, the method comprising:

determining a temperature control error and a temperature index parameter according to the temperatures of a plurality of temperature control points in the computing equipment;

if the temperature index parameter meets the adjustment condition, determining a target control parameter according to the temperature control error and a control model;

and controlling the operation of the temperature regulating module based on the target control parameter.

2. The method of claim 1, wherein determining a target control parameter based on the temperature control error and a control model if the temperature index parameter satisfies an adjustment condition comprises:

if the temperature index parameter meets the adjustment condition, iterating the process control parameter at least once according to the temperature control error and the control model; wherein the process control parameter is configured to control the temperature adjustment module to operate for a preset duration;

And if the temperature regulation module works based on the latest process control parameter so that the temperature index parameter does not meet the regulation condition, determining the latest process control parameter as the target control parameter.

3. The method of claim 1, wherein the control model is derived based on a proportional-integral-derivative PID control strategy and a neuron learning algorithm, and wherein determining the target control parameter based on the temperature control error and the control model comprises:

determining a PID parameter adjustment amount according to the recorded historical control parameter, the recorded PID parameter historical adjustment amount used for determining the historical control parameter and the temperature control historical error and the temperature control error;

and obtaining control parameters through a neuron learning algorithm according to the configured PID parameters and the PID parameter adjustment quantity.

4. The method of claim 1, wherein determining a temperature control error based on temperatures of a plurality of temperature control points within the computing device comprises:

acquiring the temperature of each temperature control point in the computing equipment, and obtaining the temperature deviation of each temperature control point according to the temperature of each temperature control point and the expected temperature;

And carrying out weighted summation according to the weight value corresponding to each temperature control point and each temperature deviation to obtain a temperature control error.

5. The method of claim 4, further comprising, prior to performing weighted summation based on the weight value corresponding to each temperature control point and each temperature deviation to obtain a temperature control error:

setting weight values corresponding to the temperature control points according to the attribute information of the heating source modules corresponding to the temperature control points;

wherein, the attribute information of the heating source module includes: either or both of the type of heat source module and the distance between the heat source module and the temperature regulation module in the computing device.

6. The method of claim 1, wherein the temperature index parameter comprises: and calculating the obtained variance according to the temperature corresponding to each temperature control point and the expected temperature.

7. The method of claim 3, wherein the neuron learning algorithm is a single neuron learning algorithm, an input of the single neuron learning algorithm is the configured PID parameter, the PID parameter adjustment is a weight value, and an output of the single neuron learning algorithm is a control parameter;

The single neuron learning algorithm sums the proportion P parameter included in the configured PID parameter and the proportion P parameter adjustment amount in the PID parameter adjustment amount, sums the integral I parameter included in the configured PID parameter and the integral I parameter adjustment amount in the PID parameter adjustment amount, sums the differential D parameter included in the configured PID parameter and the differential D parameter adjustment amount in the PID parameter adjustment amount, and predicts the control parameter through a prediction function.

8. The method of claim 2, wherein the method further comprises:

acquiring temperature change information, wherein the temperature change information comprises temperature change values before and after controlling the operation of the temperature regulation module;

if the temperature change information meets the temperature rise condition of the computing equipment, shortening the preset duration;

or if the temperature change information meets the temperature reduction condition of the computing equipment, the preset duration is increased.

9. The method of any of claims 1-8, wherein each temperature control point within the computing device corresponds to at least one heat source module, the heat source module comprising any one or more of a central processor, a graphics processor, and an image processing unit disposed within the computing device.

10. A computing device, the computing device comprising: a processor for implementing a control method of a temperature regulation module according to any one of claims 1 to 9.