CN115875298A - Fan rotating speed prediction regulation and control method and device, computer equipment and storage medium - Google Patents
Fan rotating speed prediction regulation and control method and device, computer equipment and storage medium Download PDFInfo
- Publication number
- CN115875298A CN115875298A CN202211643002.3A CN202211643002A CN115875298A CN 115875298 A CN115875298 A CN 115875298A CN 202211643002 A CN202211643002 A CN 202211643002A CN 115875298 A CN115875298 A CN 115875298A
- Authority
- CN
- China
- Prior art keywords
- output
- fan
- control input
- model
- time
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 48
- 230000033228 biological regulation Effects 0.000 title description 37
- 238000005457 optimization Methods 0.000 claims abstract description 39
- 238000012937 correction Methods 0.000 claims abstract description 37
- 230000001105 regulatory effect Effects 0.000 claims abstract description 12
- 239000011159 matrix material Substances 0.000 claims description 20
- 238000005096 rolling process Methods 0.000 claims description 18
- 238000004590 computer program Methods 0.000 claims description 15
- 230000004044 response Effects 0.000 claims description 15
- 239000013598 vector Substances 0.000 claims description 12
- 238000005070 sampling Methods 0.000 claims description 9
- 230000006870 function Effects 0.000 claims description 8
- 238000004364 calculation method Methods 0.000 claims description 5
- 239000000126 substance Substances 0.000 claims description 5
- 238000010276 construction Methods 0.000 claims description 4
- 101150036841 minJ gene Proteins 0.000 claims description 4
- 230000001276 controlling effect Effects 0.000 claims description 3
- 238000004422 calculation algorithm Methods 0.000 description 10
- 238000010586 diagram Methods 0.000 description 8
- 238000013461 design Methods 0.000 description 4
- 230000017525 heat dissipation Effects 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- UFHFLCQGNIYNRP-UHFFFAOYSA-N Hydrogen Chemical compound [H][H] UFHFLCQGNIYNRP-UHFFFAOYSA-N 0.000 description 2
- 229910052739 hydrogen Inorganic materials 0.000 description 2
- 239000001257 hydrogen Substances 0.000 description 2
- 230000001360 synchronised effect Effects 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 238000011217 control strategy Methods 0.000 description 1
- 238000001816 cooling Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 238000009795 derivation Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000004134 energy conservation Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 230000007704 transition Effects 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Images
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Abstract
The application relates to a method and a device for predicting and regulating the rotating speed of a fan, computer equipment and a storage medium. The method comprises the following steps: the temperature of a server hardware component is used as an input variable, and the reference output y at the moment k is obtained through a reference track r (k + i) the fan acquires a control input u (k) at the moment k and outputs a rotating speed y (k); model output y from a prediction model M (k + i); predicting output y of prediction model by closed loop feedback correction using output speed y (k) of fan M (k + i) performing feedback correction to obtain a predicted output y p (k + i); compensation for predicted output y by roll optimization p (k + i) calculating an optimal control input u (k); the fan takes the optimal control input u (k) at time k and adjusts the output speed y (k). The method can control the fan rotating speed in the system, and predict the future fan rotating speed by taking the temperature of each future component as an input variable, so that the algorithmThe corresponding actual operating speed is directly predicted.
Description
Technical Field
The present application relates to the technical field of fan regulation and control of servers, and in particular, to a method and an apparatus for regulating and controlling fan rotation speed through prediction, a computer device, and a storage medium.
Background
The fan control of the server adopts a model prediction algorithm to control the rotating speed of the fan in the system according to the actual running state of the component. The traditional control strategy is mainly linear control or PID (proportional integral and derivative) regulation, the linear control can cause the rotating speed to be higher, the test and verification are time-consuming and labor-consuming, the PID mode can cause the risk of the overshoot of the rotating speed of the fan, so that the temperature is lower or over-temperature is caused, the overshoot can increase the heat dissipation risk of the system, and the power consumption of the fan of the system is increased, so that the power consumption of the whole machine is increased. In the PID regulation and control mode adopted in actual operation at present, a designer in the early stage needs to set and verify different parameters for each type of machine, so certain manpower is occupied, all the parameters cannot be exhausted due to the influence of project time, and therefore each set parameter is not necessarily the optimal parameter of the project, and certain heat dissipation risk exists.
With the development of novel technologies such as cloud computing and big data, the requirements on the bandwidth and capacity of data storage are higher and higher, the operation speed and the operation amount of a processor are also higher and higher, the temperature of each component such as a memory and a hard disk is increased rapidly, the heat dissipation of an electronic device becomes a very causful problem at present, the requirements on power consumption of the society at present are lower and lower, and energy conservation is a mainstream trend at present. How to effectively reduce the overhigh temperature of each electronic component and reduce the power consumption of the fan wall, and meeting the optimal power consumption of the whole machine is a technical problem which needs to be solved urgently. Linear or PID regulation is now used. The first mode wastes time and energy and consumes power, and the second mode has the risk that the temperature of the fan and the components is overshot, so that the integral rotating speed is increased, and the power consumption of the whole machine is increased. This will cause electrical losses, reduce the PUE (power usage efficiency) value of the system, and increase the operating cost.
Disclosure of Invention
Therefore, it is necessary to provide a method, an apparatus, a computer device and a storage medium for predictive regulation of a fan rotation speed to solve the technical problems that time and energy are wasted and power is consumed due to a linear or PID regulation mode for cooling a server fan, and the overall rotation speed is increased and power consumption is increased due to the risk of overshoot of the fan and component temperatures.
In one aspect, a method for predictive regulation of fan speed is provided, where the method includes:
the temperature of a server hardware component is used as an input variable, and the reference output y at the moment k is obtained through a reference track r (k + i) the fan acquires a control input u (k) at the moment k and outputs a rotating speed y (k);
obtaining control input u (k) at time k and obtaining model output y through a prediction model M (k+i)；
Predicting output y of prediction model by closed loop feedback correction using output speed y (k) of fan M (k + i) performing feedback correction to obtain a predicted output y p (k+i)；
Compensation for predicted output y by roll optimization p (k + i) calculating an optimal control input u (k);
the fan takes the optimal control input u (k) at time k and adjusts the output speed y (k).
Further, the temperature of the server hardware component is used as an input variable, and a reference output y at the time k is obtained through a reference track r The (k + i) step includes:
starting from the actual output y (k) at time k and going to the set value y sp Establishing a reference track with excessive smoothness, wherein the reference track corresponds to expected output;
reference trajectory at time k is determined by the value y at the future time r (k + i) is expressed as:
y r (k+i)＝(1-α i )y sp +α i y(k)；
wherein, y r (k + i) is a reference output; alpha is alpha 1 Is a softening coefficient, and 0 < alpha 1 ＜1，1≤i≤p。
Further, the softening coefficient α = exp (-it/T); wherein T is the sampling period and T is the time constant of the reference trajectory.
Further, the control input u (k) at the moment k is obtained, and the model output y is obtained through a prediction model M The (k + i) step includes:
acquiring control input u (k + i-j) corresponding to the temperature of the server hardware component and sampling values g1, g2, g3.. Gn of unit impulse response, and calculating by using a discrete convolution formula to obtain model outputy M (k+i)；
The discrete convolution formula is:
wherein, y M (k + i) represents the model prediction output of the system; u (k + i-j) represents a control input of the system; g = { G 1 ，g 2 ，...g n Denotes the model vector of the system, 1 ≦ j ≦ n.
Further, the predicted output y of the prediction model is corrected by closed loop feedback using the output speed y (k) of the fan M (k + i) performing feedback correction to obtain a predicted output y p The (k + i) step includes:
the closed loop prediction output at time k is noted as y P (k)＝y M (k)+h*e(k)；
Wherein, y P (k)＝[y P (k+1)，...，y P (k+P)] T ；
h＝[h 1 ，...，h P ] T H is a feedback coefficient matrix;
further, the compensating of the predicted output y by the rolling optimization p (k + i), the formula for calculating the optimal control input u (k) includes:
wherein, P is an optimized time domain; m is a control time domain, and M is less than P; qi is an output tracking weighting coefficient; r is a radical of hydrogen j Inputting a weighting coefficient; y is p (k + i) is the prediction output; minJ (k) is the minimum value of the calculation formula.
Further, the over-roll optimization compensates for the predicted output y p (k + i) calculating an optimal control input u (k), comprising:
and solving the optimized control input under the rolling optimization performance index according to the prediction model, the reference track and the closed-loop prediction of feedback correction:
construction of
Wherein Q = diag (Q, Q) 2 ，...，q P )，R＝diag(r 1 ，r 2 ，...，r M )；G 1 Is relative to G 2 Model vector of previous time u 1 (k) Is relative u 2 (k) Control input at the previous moment;
(diag function used to construct a diagonal matrix in FreeMatlab, function that is not a square matrix with all 0 elements on the diagonal, or returns diagonal elements on a matrix in vector form.)
The solved optimal control inputs are: u (k) = d T *[y r (k)-G 2 u 2 (k)-he(k)]；
Wherein the content of the first and second substances,
/>
in another aspect, a model predictive regulation apparatus is provided, the apparatus including:
a reference track module for calculating a reference output y at the time k via a reference track by using the temperature of the server hardware component as an input variable r (k + i) the fan acquires a control input u (k) at the moment k and outputs a rotating speed y (k);
a prediction model module for obtaining control input u (k) at time k and obtaining model output y through the prediction model M (k+i)；
A feedback correction module for correcting the predicted output y of the prediction model by closed-loop feedback using the output speed y (k) of the fan M (k + i) performing feedback correction to obtain a predicted output y p (k+i)；
A rolling optimization module for compensating the predicted output y by rolling optimization p (k + i) calculating an optimal control input u (k);
and the controlled object adjusting module is used for acquiring the optimal control input u (k) at the moment k by the fan and adjusting the output rotating speed y (k).
In another aspect, a computer device is provided, which includes a memory, a processor, and a computer program stored on the memory and executable on the processor, and the processor implements the following steps when executing the computer program:
the temperature of a server hardware component is used as an input variable, and the reference output y at the moment k is obtained through a reference track r (k + i) the fan acquires a control input u (k) at the moment k and outputs a rotating speed y (k);
obtaining control input u (k) at time k and obtaining model output y through a prediction model M (k+i)；
Prediction of the output y of a predictive model using the output speed y (k) of the fan by closed loop feedback correction M (k + i) performing feedback correction to obtain a predicted output y p (k+i)；
Compensation for predicted output y by roll optimization p (k + i) calculating an optimal control input u (k);
the fan takes the optimal control input u (k) at time k and adjusts the output speed y (k).
In yet another aspect, a computer-readable storage medium is provided, having stored thereon a computer program which, when executed by a processor, performs the steps of:
taking the temperature of a server hardware component as an input variable, and obtaining a reference output y at the moment k through a reference track r (k + i) the fan acquires a control input u (k) at the moment k and outputs a rotating speed y (k);
obtaining control input u (k) at time k and obtaining model output y through a prediction model M (k+i)；
Predicting output y of prediction model by closed loop feedback correction using output speed y (k) of fan M (k + i) performing feedback correction to obtain a predicted output y p (k+i)；
Compensation for predicted output y by roll optimization p (k + i) calculating an optimal control input u (k);
the fan takes the optimal control input u (k) at time k and adjusts the output speed y (k).
According to the method, the device, the computer equipment and the storage medium for predicting and regulating the fan rotating speed, the fan regulation can be automatically and reasonably regulated and controlled under the conditions of low load or high load by adopting a model prediction method, so that the quick response to the matched fan rotating speed under different boundary conditions is ensured, the overshoot phenomenon is avoided, and the disadvantages of the conventional PID fan regulation and control are avoided. This kind of regulation and control scheme of model prediction mode can effectively solve this problem of overshooting, and the running state according to the model predecessor is as basic data, carries out automatic learning and rectifies, sets up out reasonable orbit for the fan operation in later stage, and this kind of regulation and control sets up more rationally, and the utensil is pointed, and the intellectuality promotes greatly, uses manpower sparingly, increases system security, has also saved the system consumption.
The method and the device can control the fan rotating speed in the system, and predict the future fan rotating speed by taking the temperature of each future component as an input variable, so that the corresponding actual operating speed is directly predicted through the algorithm. The method can be generally applied to general machine types, and achieves the purposes of reducing power consumption and maintaining the temperature of components to be relatively stable. After the method is applied, the universal machine type is efficiently and quickly regulated and controlled under different working conditions, the rotating speed of the fan is effectively reduced under the condition of ensuring enough air flow through the algorithm, the temperature of each component is ensured to be within a required threshold value, the power consumption of the whole machine is reduced, the noise is reduced, and the optimized design is met. The intelligent fan speed regulation system has the advantages of intelligentized speed regulation, strong regulation strategy applicability and more reasonability, thereby showing fast response and stable response, greatly saving the power consumption of the whole machine which runs for a long time and reducing the noise of the fan.
Drawings
In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings needed to be used in the description of the embodiments will be briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.
FIG. 1 is a block diagram of open loop and closed loop predictions for a fan speed prediction regulation method in one embodiment;
FIG. 2 is a diagram illustrating a model vector constructed reference trajectory of a fan speed prediction regulation method in one embodiment;
FIG. 3 is a flow chart illustrating a method for predictive regulation of fan speed according to one embodiment;
FIG. 4 is a block diagram showing the structure of a model predictive control apparatus according to an embodiment;
FIG. 5 is a diagram of the internal structure of a computer device in one embodiment.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, the present application is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present application and are not intended to limit the present application.
As described in the background art, in order to avoid the problem of temperature overshoot during the regulation and control operation of a device, the rotating speed of the fan is controlled to be extremely high, and the power consumption of the system fan is increased, so that the power consumption of the whole machine is increased. The fan resources are reasonably distributed under various loads, normal operation under different boundary conditions is guaranteed, effective and safe operation of devices of the same type is basically achieved, the disadvantages of regulation and control of the existing fan are avoided, and the local temperature of the devices corresponding to the position under the boundary condition of lower actual operation wind speed is too high, so that the rotating speed of the fan is rapidly increased.
Example 1
In order to solve the above problems, embodiment 1 of the present invention creatively provides a method for predicting and controlling a fan rotation speed, which not only can effectively and reasonably allocate fan resources and ensure sufficient air flow, but also can effectively reduce the power consumption of a whole system, and can achieve self-regulation of a machine according to different operating environments and working pressures.
The MAC is used as a nonparametric model and takes the impulse response of the system as an internal prediction model, and the future output state of the system is predicted according to the prediction model of the system through the past and current input and output states. A block diagram of open-loop prediction and closed-loop prediction is shown in fig. 1.
The basic algorithm of the fan rotating speed prediction regulation and control method comprises Model Algorithm Control (MAC), and comprises four parts: prediction model, feedback correction, reference trajectory and roll optimization.
As can be seen from the above block diagram, the MAC algorithm mainly consists of four modules in the diagram, and the meaning of each module is as follows. When the expected input is c, the control input u (k) at the moment k is obtained, and the reference output y at the moment is obtained through the reference track r (k + i) and finding the model output y from the predictive model M (k + i) obtaining the predicted output y by closed loop feedback correction p (k + i), and finally, the optimal control rate u (k) at that time is calculated by roll optimization. Wherein the feedback correction instantaneously corrects the model predicted output y with the detected output error e (k) M . The rolling optimization is to output the corrected prediction y p (k + i) and reference signal y r And (k + i) comparing, and calculating the controlled variable u (k) under various constraint conditions (such as controlled variable, output and the like) to minimize the predicted output error of the future finite time domain.
The predictive model predicts the system future response using historical information and future inputs from the controlled object based on a dynamic model of the process. And a rolling type limited time domain optimization strategy is adopted, and optimization calculation and rolling implementation are repeatedly performed on line, so that uncertainty caused by model mismatch, time variation, interference and the like can be timely compensated, and the control effect of the system is improved.
1. Prediction model
Corresponds to y M And (k + i) obtaining. For a linear system, if the sampled values g1, g2, g3... Gn of its unit impulse response are known, then using the discrete convolution formula, the prediction model description of the system can be approximated by a finite term convolution:
wherein y is M (k + i) represents the model prediction output of the system; u (k + i-j) represents a control input of the system; as shown in fig. 2, G = { G1, G2.. Gn } represents a model vector of the system, T is a sampling period, and T is a time constant of the reference trajectory, and equation (1) can be obtained by a system identification method.
2. Reference track
Corresponds to y r And (k + i) obtaining. As shown in FIG. 2, in MAC, the desired output of the control system is set to a set value y starting from the current actual output y (k) sp A reference track with a smooth transition. The reference trajectory at time k may be determined by the value y at the future time r (k + i), where i =1,2. It is characterized as follows:
y r (k+i)＝(1-α i )y sp +α i y(k)---(2)
wherein, y r (k + i) is a reference output; alpha is alpha 1 Is a softening coefficient, and 0 < alpha 1 ＜1。
α = exp (-it/T), the larger the time constant T of the reference trajectory, i.e., the larger the α value, the stronger the robustness, but the rapidity of the control is deteriorated; on the contrary, the faster the reference track reaches the set value, and the robustness is poor; therefore, in the design of MAC, α is a very important parameter, which plays an important role in the performance of the closed loop system.
3. Roll optimization
Corresponding to the determination of u (k + i). With optimal control, the roll optimization problem of Model Predictive Control (MPC) is actually to obtain the optimal control input u (k) with certain optimization criteria. Without loss of generality, the following optimization criteria are employed:
wherein, P is an optimized time domain; m is a control time domain, and M is more than P generally; q. q.s i Tracking the weighting coefficients for the output; r is j Inputting a weighting coefficient; y is p (k + i) is the prediction output.
It will be readily appreciated that to obtain the optimum control input u (k) according to the above-described optimization criteria, the predicted output y must also be known p (k + i). Thus, for y p The derivation of (k + i) introduces a third element of Model Predictive Control (MPC) "feedback correction".
4. Feedback correction
Corresponds to y p And (k + i) obtaining. Taking closed-loop prediction as an example, the closed-loop prediction output of the k-time system can be recorded as
y P (k)＝y M (k)+h*e(k)---(4)
Wherein, y P (k)＝[y P (k+1)，...，y P (k+P)] T ；
h＝[h 1 ，...，h P ] T H is a feedback coefficient matrix;
5. optimum rate of control
From the foregoing analysis, the unconstrained MAC optimization control rate under the rolling optimization performance index can be solved according to the prediction model, the reference trajectory, and the closed-loop prediction of the feedback correction:
wherein Q = diag (Q, Q) 2 ，...，q P )，R＝diag(r 1 ，r 2 ，...，r M ). The diag function is used in FreeMatlab, freMat to construct a diagonal matrix, a square matrix with all 0 elements off the diagonal, or to return diagonal elements of a matrix in the form of vectors.
The optimal instant control quantity is as follows:
u(k)＝d T *[y r (k)-G 2 u 2 (k)-he(k)]---(7)；
wherein the content of the first and second substances,
example 2
Based on the same inventive concept as embodiment 1, as shown in fig. 3, in embodiment 2, a method for predictive regulation and control of a fan rotation speed is provided, which includes the following steps:
step S1, taking the temperature of a server hardware component as an input variable, and obtaining a reference output y at the moment k through a reference track r (k + i) the fan acquires a control input u (k) at the moment k and outputs a rotating speed y (k);
s2, obtaining control input u (k) at the moment k and obtaining model output y through a prediction model M (k+i)；
Step S3, predicting output y of the prediction model by using output rotating speed y (k) of the fan through closed-loop feedback correction M (k + i) performing feedback correction to obtain a predicted output y p (k+i)；
Step S4, output y is predicted through rolling optimization compensation p (k + i) calculating an optimal control input u (k);
and S5, acquiring the optimal control input u (k) at the moment k by the fan and adjusting the output rotating speed y (k).
In this embodiment, the reference output y at the time k is obtained through a reference trajectory by using the temperature of the server hardware component as an input variable r The (k + i) step includes:
starting from the actual output y (k) at time k and going to the set value y sp Establishing a reference track with excessive smoothness, wherein the reference track corresponds to expected output;
reference trajectory at time k is determined by the value y at the future time r (k + i) is expressed as:
y r (k+i)＝(1-α i )y sp +α i y(k)；
wherein, y r (k + i) is a reference output; alpha is alpha i Is a softening coefficient, and 0 < alpha i ＜1，1≤i≤p。
In this embodiment, the softening coefficient α = exp (-it/T); wherein T is the sampling period and T is the time constant of the reference trajectory.
In this embodiment, the control input u (k) at time k is obtained and the model output y is obtained through a prediction model M The (k + i) step includes:
acquiring control input u (k + i-j) corresponding to the temperature of the server hardware component and sampling values g1, g2, g3.. Gn of unit impulse response, and calculating by using a discrete convolution formula to obtain model output y M (k+i)；
The discrete convolution formula is:
wherein, y M (k + i) represents the model prediction output of the system; u (k + i-j) represents a control input of the system; g = { G 1 ，g 2 ，...g n J is more than or equal to 1 and less than or equal to n.
In this embodiment, the predicted output y of the prediction model is corrected by closed-loop feedback using the output speed y (k) of the fan M (k + i) performing feedback correction to obtain a predicted output y p The (k + i) step includes:
the closed loop prediction output at time k is noted as y P (k)＝y M (k)+h*e(k)；
Wherein, y P (k)＝[y P (k+1)，...，y P (k+P)] T ；
h＝[h 1 ，...，h P ] T H is a feedback coefficient matrix;
in this embodiment, the predicted output y is compensated for by roll optimization p (k + i), the formula for calculating the optimal control input u (k + i) includes:
wherein P is an optimized time domain(ii) a M is a control time domain, and M is less than P; q. q.s i Tracking the weighting coefficients for the output; r is j Inputting a weighting coefficient; y is p (k + i) is the prediction output; minJ (k) is the minimum value of a calculation formula.
In this embodiment, the over-roll optimization compensates the predicted output y p (k + i) calculating an optimal control input u (k), comprising:
and solving the optimized control input under the rolling optimization performance index according to the prediction model, the reference track and the closed-loop prediction of feedback correction:
construction of
Wherein Q = diag (Q, Q) 2 ，...，q P )，R＝diag(r 1 ，r 2 ，...，r M )；G 1 Is relative to G 2 Model vector of previous time u 1 (k) Is relative u 2 (k) A control input at a previous time;
the solved optimal control inputs are: u (k) = d T *[y r (k)-G 2 u 2 (k)-he(k)]；
Wherein the content of the first and second substances,
the diag function is used in FreeMatlab, freMat to construct a diagonal matrix, a square matrix with all 0 elements off the diagonal, or to return diagonal elements of a matrix in the form of vectors.
In the method for predicting and regulating the rotating speed of the fan, the fan can be automatically and reasonably regulated and controlled under the condition of low load or high load by adopting a model prediction method, the quick response to the matched rotating speed of the fan under different boundary conditions is ensured, the overshoot phenomenon is avoided, and the disadvantages of the conventional PID fan regulation and control are avoided. This kind of regulation and control scheme of model prediction mode can effectively solve this problem of overshooting, and the running state according to the model predecessor is as basic data, carries out automatic learning and rectifies, sets up out reasonable orbit for the fan operation in later stage, and this kind of regulation and control sets up more rationally, and the utensil is pointed, and the intellectuality promotes greatly, uses manpower sparingly, increases system security, has also saved the system consumption.
The method and the device can control the fan rotating speed in the system, and predict the future fan rotating speed by taking the temperature of each future component as an input variable, so that the corresponding actual operating speed is directly predicted through the algorithm. The method can be generally applied to general machine types, and achieves the purposes of reducing power consumption and maintaining the temperature of components to be relatively stable. After the method is applied, the universal machine type is efficiently and quickly regulated and controlled under different working conditions, the rotating speed of the fan is effectively reduced under the condition of ensuring enough air flow through the algorithm, the temperature of each component is ensured to be within a required threshold value, the power consumption of the whole machine is reduced, the noise is reduced, and the optimized design is met. The intelligent fan speed regulation system has the advantages of intelligentized speed regulation, strong regulation strategy applicability and more reasonability, thereby showing fast response and stable response, greatly saving the power consumption of the whole machine which runs for a long time and reducing the noise of the fan.
It should be understood that, although the steps in the flowchart of fig. 3 are shown in order as indicated by the arrows, the steps are not necessarily performed in order as indicated by the arrows. The steps are not performed in the exact order shown and described, and may be performed in other orders, unless explicitly stated otherwise. Moreover, at least some of the steps in fig. 3 may include multiple sub-steps or multiple stages that are not necessarily performed at the same time, but may be performed at different times, and the order of performing the sub-steps or stages is not necessarily sequential, but may be performed alternately or alternately with other steps or at least some of the sub-steps or stages of other steps.
In one embodiment, as shown in FIG. 4, there is provided a model predictive control device 10 comprising: the device comprises a reference track module 1, a prediction model module 2, a feedback correction module 3, a rolling optimization module 4 and a controlled object adjusting module 5.
The reference track module 1 is used for connecting the hardware components of the serverIs used as an input variable, and a reference output y at the time k is obtained through a reference track r (k + i) the fan acquires a control input u (k) at the moment k and outputs a rotating speed y (k);
the prediction model module 2 is used for obtaining a control input u (k) at the moment k and obtaining a model output y through a prediction model M (k+i)；
The feedback correction module 3 is used for correcting the predicted output y of the prediction model by closed-loop feedback by using the output rotating speed y (k) of the fan M (k + i) performing feedback correction to obtain a predicted output y p (k+i)；
The rolling optimization module 4 is used for compensating the predicted output y through rolling optimization p (k + i) calculating an optimal control input u (k);
the controlled object adjusting module 5 is used for acquiring the optimal control input u (k) at the moment k by the fan and adjusting the output rotating speed y (k).
In this embodiment, the reference output y at the time k is obtained through a reference trajectory by using the temperature of the server hardware component as an input variable r The (k + i) step includes:
starting from the actual output y (k) at time k and going to the set value y sp Establishing a reference track with excessive smoothness, wherein the reference track corresponds to expected output;
reference trajectory at time k is determined by the value y at the future time r (k + i) is expressed as:
y r (k+i)＝(1-α i )y sp +α i y(k)；
wherein, y r (k + i) is a reference output; alpha is alpha i Is a softening coefficient, and 0 < alpha i ＜1，1≤i≤p。
In this embodiment, the softening coefficient α = exp (-it/T); wherein T is a sampling period, and T is a time constant of the reference track.
In this embodiment, the control input u (k) at time k is obtained and the model output y is obtained through a prediction model M The (k + i) step includes:
acquisition of control input u (k + i-j) corresponding to temperature of server hardware component and acquisition of unit impulse responseAnd (3) sampling g1, g2, g3... Gn, and calculating by using a discrete convolution formula to obtain a model output y M (k+i)；
The discrete convolution formula is:
wherein, y M (k + i) represents the model prediction output of the system; u (k + i-j) represents a control input of the system; g = { G 1 ，g 2 ，...g n J is more than or equal to 1 and less than or equal to n.
In this embodiment, the predicted output y of the prediction model is corrected by closed-loop feedback using the output speed y (k) of the fan M (k + i) performing feedback correction to obtain a predicted output y p The (k + i) step includes:
the closed loop prediction output at time k is noted as y P (k)＝y M (k)+h*e(k)；
Wherein, y P (k)＝[y P (k+1)，...，y P (k+P)] T ；
h＝[h 1 ，...，h P ] T H is a feedback coefficient matrix;
in this embodiment, the predicted output y is compensated for by roll optimization p (k + i), the formula for calculating the optimal control input u (k + i) includes:
wherein, P is an optimized time domain; m is a control time domain, and M is less than P; q. q.s i Tracking the weighting coefficients for the output; r is a radical of hydrogen j Inputting a weighting coefficient; y is p (k + i) is the prediction output; minJ (k) is the minimum value of a calculation formula.
In this embodiment, the over-rolling optimization compensates the predicted output y p (k + i) calculating the optimumA control input u (k) step, comprising:
and solving the optimized control input under the rolling optimization performance index according to the prediction model, the reference track and the closed-loop prediction of feedback correction:
construction of
Wherein Q = diag (Q, Q) 2 ，...，q P )，R＝diag(r 1 ，r 2 ，...，r M )；G 1 Is relative to G 2 Model vector of previous time u 1 (k) Is relative u 2 (k) A control input at a previous time;
(diag function used in FreeMatlab, function to construct a diagonal matrix, a square matrix with all 0 elements off the diagonal, or a vector to return diagonal elements on a matrix.)
The solved optimal control inputs are: u (k) = d T *[y r (k)-G 2 u 2 (k)-he(k)]；
Wherein the content of the first and second substances,
in the model prediction regulation and control device, the fan regulation and control can be automatically rationalized under the condition of low load or high load by adopting the model prediction method to regulate and control, the quick response to the matched fan rotating speed under different boundary conditions is ensured, the overshoot phenomenon is avoided, and the disadvantages of the conventional PID fan regulation and control are avoided. This kind of regulation and control scheme of model prediction mode can effectively solve this problem of overshooting, and the running state according to the model predecessor is as basic data, carries out automatic study and corrects, sets out reasonable orbit for the fan operation in later stage, and this kind of regulation and control sets up more rationally, has pertinence, and the intellectuality promotes greatly, uses manpower sparingly, increases system security, has also saved the system consumption.
The method and the device can control the fan rotating speed in the system, and predict the future fan rotating speed by taking the temperature of each future component as an input variable, so that the corresponding actual operating speed is directly predicted through the algorithm. The method can be generally applied to general machine types, and achieves the purposes of reducing power consumption and maintaining the temperature of components to be relatively stable. After the method is applied, the universal machine type is efficiently and quickly regulated and controlled under different working conditions, the rotating speed of the fan is effectively reduced under the condition of ensuring enough air flow through the algorithm, the temperature of each component is ensured to be within a required threshold value, the power consumption of the whole machine is reduced, the noise is reduced, and the optimized design is met. The intelligent fan control system has the advantages of intelligent rotation speed regulation, strong regulation strategy applicability and reasonability, so that the response is fast, the response is stable, the power consumption of the whole machine which runs for a long time is greatly saved, and the noise of the fan is reduced.
For the specific definition of the model predictive control device, reference may be made to the above definition of the fan speed predictive control method, which is not described herein again. The modules in the model predictive control device can be wholly or partially realized by software, hardware and a combination thereof. The modules can be embedded in a hardware form or independent from a processor in the computer device, and can also be stored in a memory in the computer device in a software form, so that the processor can call and execute operations corresponding to the modules.
In one embodiment, a computer device is provided, which may be a server, the internal structure of which may be as shown in fig. 5. The computer device includes a processor, a memory, a network interface, and a database connected by a system bus. Wherein the processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device comprises a nonvolatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of an operating system and computer programs in the non-volatile storage medium. The database of the computer device is used for storing model predictive regulatory data. The network interface of the computer device is used for communicating with an external terminal through a network connection. The computer program is executed by a processor to realize a method for predicting and regulating the rotating speed of the fan.
Those skilled in the art will appreciate that the architecture shown in fig. 5 is merely a block diagram of some of the structures associated with the disclosed aspects and is not intended to limit the computing devices to which the disclosed aspects apply, as particular computing devices may include more or less components than those shown, or may combine certain components, or have a different arrangement of components.
In one embodiment, there is provided a computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the following steps when executing the computer program:
the temperature of a server hardware component is used as an input variable, and the reference output y at the moment k is obtained through a reference track r (k + i) the fan acquires a control input u (k) at the moment k and outputs a rotating speed y (k);
obtaining control input u (k) at time k and obtaining model output y through a prediction model M (k+i)；
Predicting output y of prediction model by closed loop feedback correction using output speed y (k) of fan M (k + i) performing feedback correction to obtain a predicted output y p (k+i)；
Compensation of predicted output y by roll optimization p (k + i) calculating an optimal control input u (k);
the fan takes the optimal control input u (k) at time k and adjusts the output speed y (k).
The specific limitations regarding the implementation steps when the processor executes the computer program can be referred to the limitations of the method for model predictive control described above, and will not be described herein again.
In one embodiment, a computer-readable storage medium is provided, having a computer program stored thereon, which when executed by a processor, performs the steps of:
the temperature of a server hardware component is used as an input variable, and the reference output y at the moment k is obtained through a reference track r (k + i) the fan acquires a control input u (k) at the moment k and outputs a rotating speed y (k);
control input for acquiring k timeEntering u (k) and finding the model output y from the prediction model M (k+i)；
Predicting output y of prediction model by closed loop feedback correction using output speed y (k) of fan M (k + i) performing feedback correction to obtain a predicted output y p (k+i)；
Compensation for predicted output y by roll optimization p (k + i) calculating an optimal control input u (k);
the fan takes the optimal control input u (k) at time k and adjusts the output speed y (k).
For specific limitations of the implementation steps when the computer program is executed by the processor, reference may be made to the above limitations of the method for model predictive control, which are not described herein again.
It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above may be implemented by hardware instructions of a computer program, which may be stored in a non-volatile computer-readable storage medium, and when executed, may include the processes of the embodiments of the methods described above. Any reference to memory, storage, database, or other medium used in the embodiments provided herein may include non-volatile and/or volatile memory, among others. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically Programmable ROM (EPROM), electrically Erasable Programmable ROM (EEPROM), or flash memory. Volatile memory can include Random Access Memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms such as Static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double Data Rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous Link DRAM (SLDRAM), rambus (Rambus) direct RAM (RDRAM), direct Rambus Dynamic RAM (DRDRAM), and Rambus Dynamic RAM (RDRAM), among others.
All possible combinations of the technical features in the above embodiments may not be described for the sake of brevity, but should be considered as being within the scope of the present disclosure as long as there is no contradiction between the combinations of the technical features.
The above-mentioned embodiments only express several embodiments of the present application, and the description thereof is more specific and detailed, but not construed as limiting the scope of the invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the concept of the present application, which falls within the scope of protection of the present application. Therefore, the protection scope of the present patent shall be subject to the appended claims.
Claims (10)
1. A method for predicting and regulating the rotating speed of a fan is used for controlling the rotating speed of the fan of a server, and is characterized by comprising the following steps:
the temperature of a server hardware component is used as an input variable, and the reference output y at the moment k is obtained through a reference track r (k + i) the fan acquires a control input u (k) at the moment k and outputs a rotating speed y (k);
obtaining control input u (k) at time k and obtaining model output y through a prediction model M (k+i)；
Prediction of the output y of a predictive model using the output speed y (k) of the fan by closed loop feedback correction M (k + i) performing feedback correction to obtain a predicted output y p (k+i)；
Compensation of predicted output y by roll optimization p (k + i) calculating an optimal control input u (k);
the fan takes the optimal control input u (k) at time k and adjusts the output speed y (k).
2. The method as claimed in claim 1, wherein the reference output y at time k is obtained from the reference trajectory by using the temperature of the server hardware component as an input variable r The (k + i) step includes:
starting from the actual output y (k) at time k and going to the set value y sp Establishing a reference track with excessive smoothness, wherein the reference track corresponds to expected output;
at time kReference trajectory by value y at future time r (k + i) is expressed as:
y r (k+i)＝(1-α i )y sp +α i y(k)；
wherein, y r (k + i) is a reference output; alpha is alpha i Is a coefficient of softening, and 0<α i <1，1≤i≤p。
3. The method of claim 2, wherein the softening coefficient α = exp (-it/T); wherein T is the sampling period and T is the time constant of the reference trajectory.
4. The method as claimed in claim 3, wherein the obtaining of the control input u (k) at time k and the obtaining of the model output y from the predictive model are performed M The (k + i) step includes:
acquiring control input u (k + i-j) corresponding to the temperature of the server hardware component and sampling values g1, g2, g3 … gn of unit impulse response, and calculating by using a discrete convolution formula to obtain model output y M (k+i)；
The discrete convolution formula is:
wherein, y M (k + i) represents the model prediction output of the system; u (k + i-j) represents a control input of the system; g = { G 1 ,g 2 ,…g n J is more than or equal to 1 and less than or equal to n.
5. The method as claimed in claim 4, wherein the predicted output y of the prediction model is corrected by closed-loop feedback using the output speed y (k) of the fan M (k + i) performing feedback correction to obtain a predicted output y p The (k + i) step includes:
the closed loop prediction output at time k is noted as y P (k)＝yM(k)+h*e(k)；
Wherein, y P (k)＝[y P (k+1)，...，y P (k+P)] T ；
h＝[h 1 ，...，h P ] T H is a feedback coefficient matrix;
6. the predictive method of claim 5, wherein the predicted output y is compensated for by roll optimization p (k + i), the formula for calculating the optimal control input u (k + i) includes:
wherein, P is an optimized time domain; m is a control time domain, and M is less than P; q. q of i Tracking the weighting coefficients for the output; r is j Inputting a weighting coefficient; y is p (k + i) is the prediction output; minJ (k) is the minimum value of a calculation formula.
7. The method as claimed in claim 6, wherein the over-roll optimization compensates the predicted output y p (k + i) calculating an optimal control input u (k), comprising:
and solving the optimized control input under the rolling optimization performance index according to the prediction model, the reference track and the closed-loop prediction of feedback correction:
construction of
Wherein Q = diag (Q, Q) 2 ，...，q P )，R＝diag(r 1 ，r 2 ，...，r M )；G 1 Is relative to G 2 Model vector of previous time u 1 (k) Is relative u 2 (k) Control input at the previous moment;
(diag function used to construct a diagonal matrix in FreeMatlab, function that is not a square matrix with all 0 elements on the diagonal, or returns diagonal elements on a matrix in vector form.)
The solved optimal control inputs are: u (k) = d T *[y r (k)-G 2 u 2 (k)-he(k)]；
Wherein the content of the first and second substances,
8. a model predictive control apparatus, the apparatus comprising:
the reference track module is used for taking the temperature of the server hardware component as an input variable, solving a reference output yr (k + i) at the moment k through a reference track, and acquiring a control input u (k) at the moment k and outputting a rotating speed y (k) by a fan;
a prediction model module for obtaining control input u (k) at time k and obtaining model output y through the prediction model M (k+i)；
A feedback correction module for correcting the predicted output y of the prediction model by closed-loop feedback using the output speed y (k) of the fan M (k + i) performing feedback correction to obtain a predicted output y p (k+i)；
A rolling optimization module for compensating the predicted output y by rolling optimization p (k + i) calculating an optimal control input u (k);
and the controlled object adjusting module is used for acquiring the optimal control input u (k) at the moment k by the fan and adjusting the output rotating speed y (k).
9. A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the steps of the method of any of claims 1 to 7 are implemented when the computer program is executed by the processor.
10. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the method of any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211643002.3A CN115875298A (en) | 2022-12-20 | 2022-12-20 | Fan rotating speed prediction regulation and control method and device, computer equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211643002.3A CN115875298A (en) | 2022-12-20 | 2022-12-20 | Fan rotating speed prediction regulation and control method and device, computer equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115875298A true CN115875298A (en) | 2023-03-31 |
Family
ID=85754148
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211643002.3A Pending CN115875298A (en) | 2022-12-20 | 2022-12-20 | Fan rotating speed prediction regulation and control method and device, computer equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115875298A (en) |
-
2022
- 2022-12-20 CN CN202211643002.3A patent/CN115875298A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20070061021A1 (en) | Combining power prediction and optimal control approaches for performance optimization in thermally limited designs | |
| WO2016195117A1 (en) | Method and system for controlling operation of machine | |
| JP2010514986A (en) | Method for computer-aided closed-loop control and / or open-loop control of technical systems, in particular gas turbines | |
| CN107270283B (en) | Multivariable constraint predictive control method based on circulating fluidized bed unit | |
| JP6901037B1 (en) | Control devices, control methods and programs | |
| CN116301275B (en) | Energy consumption adjusting method and device, electronic equipment and medium | |
| CN108241526B (en) | Adaptive control strategy based on mode prediction | |
| Baranski et al. | Comparative study of neighbor communication approaches for distributed model predictive control in building energy systems | |
| CN116126052A (en) | Method, apparatus, device and storage medium for temperature control | |
| Gruber et al. | Min–Max MPC based on an upper bound of the worst case cost with guaranteed stability. Application to a pilot plant | |
| JP7014330B1 (en) | Controls, control methods, and programs | |
| Naug et al. | A relearning approach to reinforcement learning for control of smart buildings | |
| CN115875298A (en) | Fan rotating speed prediction regulation and control method and device, computer equipment and storage medium | |
| Krishnamoorthy et al. | Adaptive horizon economic nonlinear model predictive control | |
| CN111197840B (en) | Control method and device of combined type wind cabinet, computer equipment and storage medium | |
| CN110454322B (en) | Water turbine speed regulation control method, device and system based on multivariable dynamic matrix | |
| Zandi et al. | Voltage control of a quasi z-source converter under constant power load condition using reinforcement learning | |
| CN112306122A (en) | Greenhouse environment control method, device, system, computer equipment and storage medium | |
| Lee et al. | Value function-based approach to the scheduling of multiple controllers | |
| JPH11119805A (en) | Correction method of estimated amount of demanded electric power | |
| Coffman et al. | A study of virtual energy storage from thermostatically controlled loads under time-varying weather conditions | |
| Karimi‐Pour et al. | Economic model predictive control of nonlinear systems using a linear parameter varying approach | |
| CN115494930A (en) | Optimal low energy consumption keeping method, device, equipment and medium | |
| CN112948644A (en) | Valve opening control method and device, computer equipment and storage medium | |
| Sheikhi et al. | Design of nonlinear predictive generalized minimum variance control for performance monitoring of nonlinear control systems |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |