Heat dissipation control method

Abstract

The application discloses a heat dissipation control method, relates to the technical field of heat dissipation control, and comprises the following steps: collecting an actual temperature in the operation of an electronic device, calculating a temperature error according to an expected temperature and the actual temperature of the electronic device, and calculating a temperature error change rate according to the temperature error; outputting a first adjustment amount of a heat dissipation object by using a pre-constructed fuzzy controller, and adjusting to make the temperature be in a certain stable interval; generating a second adjustment amount of the heat dissipation object according to the temperature and power consumption, and adjusting to make the power consumption satisfy certain optimization conditions, so that the problems that a PID controller faces response lag, large steady-state error and the like in a nonlinear temperature system are solved, the chattering effect exists in sliding mode control, the fan mechanical loss is easily caused, and then the service life and reliability are reduced, technical problems are solved, the chattering effect is eliminated, the control smoothness is improved, high-frequency switching is avoided, the fan mechanical loss is reduced, the state estimation error is reduced, the response speed is improved, and the response time is shortened.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of heat dissipation control, and in particular to a heat dissipation control method. BACKGROUND

[0002] In the related art, the temperature of an electronic device (such as a switch, a storage device, a server, etc.) can be controlled by a proportional integral derivative controller (PID controller for short), that is, a real-time response to a current temperature error (such as a difference between a set value and an actual value) is performed through proportional control, and then the fan speed is adjusted; long-term deviation is eliminated through integral control; the temperature change trend is predicted through differential control to suppress temperature fluctuations, and thus the temperature of the electronic device is stabilized near the target value; or the heat dissipation of the electronic device can be achieved through sliding mode control, that is, the state of the electronic device is made to slide along a preset sliding surface through the design of the sliding surface, and finally the temperature of the electronic device is quickly converged to the target range.

[0003] However, in the related art, in a nonlinear temperature system, the PID controller often faces the problem of response lag, is difficult to quickly follow the dynamic changes of the system temperature, and produces a large steady-state error, resulting in a significant deviation between the actual temperature and the set temperature of the system; and the sliding mode control has a chattering effect, and this high-frequency switching control makes the fan frequently change the operating state, and thus aggravates the mechanical loss of the fan, shortens the service life, and needs to be improved. SUMMARY

[0004] The present application provides a heat dissipation control method to at least solve the problems in the related art that the PID controller faces response lag, large steady-state error, etc. in a nonlinear temperature system, the sliding mode control has a chattering effect, is easy to cause mechanical loss of the fan, and thus reduces the service life and reliability of the fan, etc.

[0005] The present application provides a heat dissipation control method applied to an electronic device having a heat dissipation object, wherein the method comprises: collecting an actual temperature in the running of the electronic device, and calculating a corresponding temperature error based on the actual temperature and an expected temperature of the electronic device, and calculating a temperature error change rate of the electronic device according to the temperature error; inputting the temperature error and / or the temperature error change rate into a pre-constructed fuzzy controller to output a first adjustment amount of the heat dissipation object, and adjusting the heat dissipation object according to the first adjustment amount to make the temperature of the electronic device in a preset stable interval; generating a second adjustment amount of the heat dissipation object according to the temperature and power consumption of the electronic device, and adjusting the heat dissipation object according to the second adjustment amount to make the power consumption of the electronic device meet a preset optimization condition.

[0006] This application also provides a heat dissipation control device applied to an electronic device with a heat dissipation object. The device includes: a calculation module for acquiring the actual temperature of the electronic device during operation, calculating a corresponding temperature error based on the actual temperature and the desired temperature of the electronic device, and calculating the temperature error change rate of the electronic device based on the temperature error; a first adjustment module for inputting the temperature error and / or the temperature error change rate into a pre-built fuzzy controller to output a first adjustment amount for the heat dissipation object, and adjusting the heat dissipation object according to the first adjustment amount to keep the temperature of the electronic device within a preset stable range; and a second adjustment module for generating a second adjustment amount for the heat dissipation object based on the temperature and power consumption of the electronic device, and adjusting the heat dissipation object according to the second adjustment amount to ensure that the power consumption of the electronic device meets preset optimization conditions.

[0007] This application also provides a controller, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of any of the above-described heat dissipation control methods.

[0008] This application also provides an electronic device including the controller described above, the electronic device being used to implement the steps of any of the above heat dissipation control methods.

[0009] This application also provides a computer-readable storage medium storing a computer program, wherein the computer program, when executed by a processor, implements the steps of any of the above-described heat dissipation control methods.

[0010] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of any of the above-described heat dissipation control methods.

[0011] This application allows for the calculation of temperature error based on the actual and desired temperatures of an electronic device during operation. The corresponding rate of change of temperature error is then calculated, and a pre-built fuzzy controller outputs a first adjustment value for the heat dissipation object, which is then adjusted to keep the temperature of the electronic device within a certain stable range. A second adjustment value for the heat dissipation object is then generated based on the temperature and power consumption of the electronic device, and adjusted to ensure that the power consumption of the electronic device meets certain optimization conditions. Therefore, this application solves the problems of PID controllers in nonlinear temperature systems, such as response lag and large steady-state error, and addresses the chattering effect in sliding mode control, which easily leads to fan mechanical wear and reduces its service life and reliability. The application achieves the technical effects of eliminating chattering, improving control smoothness, avoiding high-frequency switching, reducing fan mechanical wear, reducing state estimation error, improving response speed, and shortening response time. Attached Figure Description

[0012] In order to more clearly illustrate the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiments. Obviously, the drawings described in the following embodiments are only some of the embodiments of the present application, and other drawings can be obtained by those skilled in the art without any creative effort based on these drawings.

[0013] Figure 1 A flow chart of a heat dissipation control method according to an embodiment of the present application;

[0014] Figure 2 A particle filtering flow chart according to an embodiment of the present application;

[0015] Figure 3 A back propagation neural network prediction flow chart according to an embodiment of the present application;

[0016] Figure 4 A fuzzy controller working flow chart according to an embodiment of the present application;

[0017] Figure 5 A working principle flow chart of a heat dissipation control method according to an embodiment of the present application;

[0018] Figure 6 A block schematic diagram of a heat dissipation control device according to an embodiment of the present application.

[0019] Reference signs:

[0020] Wherein, 10-heat dissipation control device; 100-first calculation module, 200-first adjustment module, 300-second adjustment module. DETAILED DESCRIPTION

[0021] The technical solutions in the embodiments of the present application will be described clearly and completely in the following with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, but not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without any creative effort belong to the protection scope of the present application.

[0022] It should be noted that, in the description of the present application, the terms “include”, “contain” or any other variants thereof are intended to cover non-exclusive inclusion, so that the process, method, article or device including a series of elements not only includes those elements, but also includes other elements not explicitly listed or inherent to such process, method, article or device. The terms “first”, “second” and the like in the present application are used to distinguish similar objects, and are not used to describe a specific order or sequence.

[0023] In order for those skilled in the art to better understand the present application, the present application will be further described in detail below in conjunction with the accompanying drawings and specific embodiments.

[0024] Embodiments of the present application provide a heat dissipation control method, which is described in detail in combination with the execution flow of the heat dissipation control method.

[0025] Specifically, Figure 1 A flowchart of a heat dissipation control method according to an embodiment of the present application is shown.

[0026] As Figure 1 shown, the heat dissipation control method is applied to an electronic device having a heat dissipation object, and the heat dissipation control method comprises the following steps:

[0027] In step S101, the actual temperature of the electronic device in operation is collected, and based on the actual temperature and the expected temperature of the electronic device, the corresponding temperature error is calculated, and the temperature error change rate of the electronic device is calculated according to the temperature error.

[0028] It can be understood that the actual temperature of the electronic device can be collected by a temperature sensor, or the actual temperature of the electronic device can be obtained by using a system command, which can be set by those skilled in the art according to the actual situation, and the present application does not make specific limitations.

[0029] In addition, the electronic device in the embodiments of the present application can be a switch, a storage device, a server, a PC (Personal Computer), and other devices having a heat dissipation object, and can also be other devices having a heat dissipation object, which can be set by those skilled in the art according to the actual situation, and the present application does not make specific limitations.

[0030] In some embodiments, the actual temperature of the electronic device in operation can be collected, the expected temperature of the electronic device can be obtained, and then the temperature error can be calculated according to the actual temperature and the corresponding expected temperature of the electronic device, so as to calculate the temperature error change rate of the electronic device according to the temperature error.

[0031] For example, after the server is powered on, the actual temperature inside the server (such as the central processing unit, the cabinet area, etc., which are not limited in the present application) can be collected by a temperature sensor, and the corresponding expected temperature can be obtained, and then the corresponding temperature error can be calculated by using the calculation formula of the temperature error, and the corresponding temperature error change rate can be calculated according to the calculation formula of the temperature error and the temperature error change rate.

[0032] The calculation formula of the temperature error can be, but is not limited to:

[0033] ,

[0034] in, The desired temperature; This refers to the actual temperature, i.e., the real temperature.

[0035] The formula for calculating the rate of change of temperature error can be, but is not limited to, the following:

[0036] ,

[0037] in, For time.

[0038] Optionally, in one embodiment of this application, collecting the actual temperature during the operation of the electronic device includes: determining the initial particle state value and initial particle weight value corresponding to different particles based on the actual temperature; calculating the predicted particle state value of different particles at different times based on the initial particle state value and initial particle weight value; determining the predicted particle weight value of different particles based on the predicted particle state value; determining the particles that meet the preset particle conditions based on the predicted particle weight value; and obtaining the final actual temperature based on the particles that meet the preset particle conditions.

[0039] It is understood that the embodiments of this application introduce particle filtering to replace Kalman filtering to adapt to the nonlinear characteristics of the temperature system. The particle filtering process designed for the temperature system is as follows: Figure 2 As shown, the main content is as follows:

[0040] Step S201: Particle initialization.

[0041] In this embodiment, the temperature can be generated near the initial temperature based on the actual temperature. (among them, The particles are uniformly distributed in weight, thus yielding the corresponding initial particle state values ​​and initial particle weight values. Their expressions can be, but are not limited to, as follows:

[0042] ,

[0043] in, Indicates the first The state values ​​of the initial particles (corresponding to the state quantities of the initial temperature error); Indicated by For the mean, The variance follows a normal distribution, and the particles are at an initial temperature. The surrounding area is generated according to this normal distribution; This represents the initial temperature value, which is the central reference temperature for particle generation; The variance representing the initial particle distribution determines the degree of particle dispersion near the initial temperature. Indicates the first The initial weights of the particles are given, and initially all particle weights are uniformly distributed, hence... ; This represents the total number of particles, which is set to 1000 here to describe the number of particles in the system state.

[0044] Step S202: State prediction.

[0045] In this application embodiment, a state transition model is constructed based on the thermal conduction hysteresis, and its expression may be, but is not limited to, as follows:

[0046] ,

[0047] in, Indicates the first Time of the first The predicted state value of each particle; Indicates the first Time of the first The state values ​​of each particle; Indicates the first Time of the first The state value of each particle is introduced to account for the hysteresis of heat conduction. This represents process noise, used to simulate random disturbances in a system. , follows a mean of 0 and a variance of normal distribution , The variance of the process noise represents the intensity of the process noise.

[0048] Step S203: Weight update.

[0049] In this embodiment, a likelihood function is constructed based on the actual temperature collected by a temperature sensor, and its expression may be, but is not limited to, as follows:

[0050] ,

[0051] in, It is an identity observation model; Indicates the first The measured value of the temperature sensor at any given time; The variance of the observation noise reflects the intensity of the noise measured by the temperature sensor.

[0052] Step S204: Determine the particles that meet certain particle conditions to obtain the final actual temperature.

[0053] Certain particle conditions can be set by those skilled in the art according to actual circumstances, and this application does not impose specific restrictions.

[0054] The embodiment of the present application can determine the initial state and weight value of the particle according to the actual temperature in the running of the electronic device, and determine the final actual temperature through multi-step calculation and screening of particles meeting certain conditions, multi-particle collaborative prediction, improve the accuracy and robustness of temperature measurement, effectively handle nonlinearity and uncertainty, more accurate than PID or linear control, recursive prediction mechanism, realize real-time tracking of temperature change, and adapt to complex environment and scalability.

[0055] Optionally, in an embodiment of the present application, based on the particle prediction weight value, the particle meeting the preset particle condition is determined, including: obtaining particle expected weight values of different particles; constructing an effective evaluation model of different particles based on the particle expected weight values; inputting the particle prediction weight value into the effective evaluation model to obtain effective evaluation values of different particles; judging whether the effective evaluation value is less than a preset threshold; if the effective evaluation value is less than the preset threshold, determining that the different particles are invalid, and re-generating the different particles to obtain new particles after re-generation, and determining the particle meeting the preset particle condition based on the new particles; if the effective evaluation value is greater than or equal to the preset threshold, determining that the different particles are valid, and determining the particle meeting the preset particle condition based on the valid particles.

[0056] In some embodiments, the embodiment of the present application can obtain particle expected weight values of different particles, and then construct an effective evaluation model of different particles, wherein the expression of the effective evaluation model can be but is not limited to:

[0057] ,

[0058] wherein, is a key index for measuring particle diversity.

[0059] Further, the embodiment of the present application inputs the particle prediction weight value into the effective evaluation model to obtain the effective evaluation values of different particles.

[0060] In some embodiments, the embodiment of the present application can judge whether the effective evaluation value is less than a certain threshold, if it is less than, determining that the different particles are invalid, and re-generating the different particles to re-determine the particle meeting the certain particle condition. Wherein the certain threshold can be set by the person skilled in the art according to the actual situation, and the present application does not make specific limitation.

[0061] For example, when the embodiment of the present application is , it indicates that the particle diversity is poor, and the particle can be re-generated by the system resampling method: eliminating particles with small weights, copying particles with high weights, and making the new particle set have uniform weight distribution again, so as to restore the diversity of the particles and ensure that the particle filter can continue to effectively estimate the system state.

[0062] In some embodiments, the embodiments of the present application can determine that different particles are valid when the effective evaluation value is greater than or equal to a certain threshold value, and then determine the particles that meet a certain particle condition.

[0063] The embodiments of the present application can obtain particle expected weight values according to particle prediction weight values and construct an effective evaluation model, and then obtain the effective evaluation values of different particles by inputting the prediction weight values, determine the particle effectiveness according to the comparison result with a certain threshold value, and then determine the particles that meet a certain particle condition, avoiding the limitations of subjective judgment or simple threshold comparison, improving the accuracy and adaptability of particle screening, enhancing the system robustness and anti-interference ability, optimizing the calculation efficiency and resource utilization, prolonging the equipment life and reducing the maintenance cost.

[0064] Optionally, in an embodiment of the present application, before the actual temperature is input into the pre-constructed back propagation neural network, it further includes: obtaining the layer number information of the back propagation neural network; obtaining the structure information of different layers of the back propagation neural network; constructing the back propagation neural network based on the layer number information and the structure information.

[0065] It can be understood that in the embodiments of the present application, the layer number information can include but is not limited to the input layer, such as the number of nodes of the input layer, etc., which is not limited by the present application; the hidden layer, such as the number of nodes and the number of layers of the hidden layer, etc., which is not limited by the present application; the output layer, such as the number of nodes and the number of layers of the output layer, etc., which is not limited by the present application.

[0066] Further, in the embodiments of the present application, the structure information of different layers can include but is not limited to weight matrix, activation function, bias vector, etc., which is not limited by the present application.

[0067] In some embodiments, the embodiments of the present application can first obtain the layer number information of the back propagation neural network, and then determine the structure information of different layers to construct the back propagation neural network.

[0068] For example, the embodiments of the present application can construct a 3-layer back propagation neural network, including an input layer, a hidden layer and an output layer. Among them, the input layer has 5 nodes, the hidden layer has 10 nodes, and the output layer has 1 node, which is not limited by the present application.

[0069] Before the actual temperature is input into the pre-constructed back propagation neural network, the embodiments of the present application first obtain the layer number and the structure information of each layer, and then construct the corresponding back propagation neural network, realize the adaptive optimization of network structure, reduce the calculation overhead, speed up the network convergence, reduce the training cost and improve the convergence speed, improve the generalization ability, improve the accuracy and robustness of temperature prediction, support transfer learning and cross-scene application, and improve the data utilization rate.

[0070] Optionally, in an embodiment of the present application, before the actual temperature is input into the pre-constructed back propagation neural network, further comprising: obtaining a training temperature corresponding to the back propagation neural network; determining a training function and a prediction function of the back propagation neural network based on the training temperature; and constructing the back propagation neural network based on the training function and the prediction function.

[0071] As a possible implementation manner, the embodiment of the present application can first obtain a training temperature corresponding to the back propagation neural network, and then determine the corresponding training function and prediction function, so as to construct the back propagation neural network.

[0072] In the embodiment of the present application, the corresponding training function can be determined based on the Levenberg-Marquardt algorithm. The Levenberg-Marquardt algorithm is an iterative optimization algorithm for solving nonlinear least squares problems.

[0073] Before the actual temperature is input into the pre-constructed back propagation neural network, the embodiment of the present application first obtains the corresponding training temperature, and then determines the training and prediction functions, so as to construct the corresponding back propagation neural network. The dynamic topology design avoids overfitting and underfitting, improves the accuracy and robustness of temperature prediction, and optimizes the calculation efficiency and resource utilization.

[0074] Optionally, in an embodiment of the present application, the prediction function of the back propagation neural network is determined based on the training temperature, comprising: determining a first weight matrix of an input layer to a hidden layer of the back propagation neural network based on the training temperature; determining a first bias vector and a first activation function of the hidden layer based on the first weight matrix; determining a second weight matrix of the hidden layer to an output layer of the back propagation neural network based on the first bias vector and the first activation function; determining a second bias vector and a second activation function of the output layer based on the second weight matrix; and constructing the prediction function based on the first weight matrix, the first bias vector, the first activation function, the second weight matrix, the second bias vector and the second activation function.

[0075] In some embodiments, when constructing the prediction function, the embodiment of the present application can first determine the first weight matrix of the input layer to the hidden layer, the first bias vector and the first activation function of the hidden layer, the second weight matrix of the hidden layer to the output layer, and the second bias vector and the second activation function of the output layer, and then obtain the corresponding prediction function. The expression of the prediction function can be, but is not limited to:

[0076] ,

[0077] wherein, is the input vector of the neural network. In combination with the foregoing, the input here is the current temperature and the temperature values at the previous four time points, i.e. , which contains historical temperature information and is used to capture the regularity of temperature changes; represents the weight matrix from the input layer to the hidden layer, which is used to measure the influence degree of each element in the input vector on the hidden layer neurons, and is continuously adjusted through training to optimize the network performance; represents the bias vector of the hidden layer, which is used to adjust the activation threshold of the hidden layer neurons, so that the network can better fit the data; represents the activation function (Sigmoid activation function can be selected, which is not specifically limited in the present application), which is used to introduce non-linear characteristics to the hidden layer neurons, so that the network can handle non-linear temperature change relationships, and map the results of linear combination to a specific range (such as between 0 and 1); represents the weight matrix from the hidden layer to the output layer, which determines the influence weight of the hidden layer output on the final prediction result; represents the bias of the output layer, which is used to adjust the activation threshold of the output layer, so that the output result is more in line with the actual temperature prediction requirements; represents the activation function of the output layer, which is selected according to the actual situation (for example, if the predicted temperature is a continuous value, a linear activation function can be used), which is used to convert the results processed by the hidden layer into the final temperature prediction value.

[0078] The embodiments of the present application can determine the first weight matrix from the input layer to the hidden layer of the back propagation neural network, the first bias vector and the activation function of the hidden layer, the second weight matrix from the hidden layer to the output layer, the second bias vector and the activation function of the output layer based on the training temperature, and then construct a prediction function, dynamically design the weight matrix, accurately capture the temperature characteristics, dynamically design the bias vector, compensate for the temperature reference deviation, dynamically design the activation function, adapt to the temperature nonlinear characteristics, construct the prediction function, and realize end-to-end optimization and efficient inference.

[0079] Optionally, in an embodiment of the present application, based on the actual temperature and the expected temperature of the electronic device, the corresponding temperature error is calculated, and the temperature error rate of the electronic device is calculated according to the temperature error, including: inputting the actual temperature into the pre-constructed back propagation neural network to obtain the predicted temperature of the electronic device; based on the actual temperature and the expected temperature, and / or the predicted temperature and the expected temperature, the temperature error is calculated; based on the temperature error, the temperature error rate is calculated.

[0080] In some embodiments, the embodiments of the present application can predict the temperature of the electronic device by using the pre-constructed back propagation neural network to obtain the corresponding predicted temperature.

[0081] For example, the input of the back propagation neural network of the embodiments of the present application can be the current temperature and the temperature at the previous 4 time points , and the output can be the predicted temperature for 10 seconds , which can be set by those skilled in the art according to actual conditions, and the present application does not make specific limitations, and the present application can be combined with Figure 3 The main content is shown as follows:

[0082] Step S301: temperature sensor.

[0083] In the embodiment of the present application, the temperature of the server can be collected in real time by using the temperature sensor.

[0084] Step S302: particle filter denoising.

[0085] In the embodiment of the present application, the particle filter denoising can be used, and then the final actual temperature is obtained. Figure 2

[0086] Step S303: back propagation neural network prediction.

[0087] In the embodiment of the present application, the back propagation neural network constructed in advance can be used for temperature prediction, and then the corresponding predicted temperature is obtained.

[0088] Step S304: fuzzy controller decision.

[0089] In the embodiment of the present application, the temperature error can be calculated based on the predicted temperature and the expected temperature, and the actual temperature and the expected temperature, and then the temperature error change rate is calculated, so as to use the corresponding fuzzy controller to make a decision.

[0090] Further, the temperature error between the actual temperature and the expected temperature, and the temperature error between the predicted temperature and the expected temperature can be calculated by using the above-mentioned temperature error calculation formula, and then the temperature error change rate is calculated by using the above-mentioned temperature error change rate calculation formula.

[0091] For example, the temperature error corresponding to the current temperature, the temperature error corresponding to the temperature at the previous four time points, and the temperature error corresponding to the predicted temperature at 10 seconds can be calculated, and then according to the temperature error corresponding to different temperatures, the temperature error change rate calculation formula is used to calculate the corresponding temperature error change rate.

[0092] In the embodiment of the present application, the actual temperature is input into the back propagation neural network constructed in advance to obtain the predicted temperature, and then the temperature error is calculated between the actual temperature and the expected temperature, so as to calculate the corresponding temperature error change rate, form a closed loop control of actual temperature, predicted temperature, error calculation and control adjustment, improve the temperature control precision, dynamically adjust the weight through continuous learning, optimize the self-adaptive ability of the model, and improve the calculation efficiency and scalability of the model.

[0093] ​Optionally, in an embodiment of the present application, before the temperature error and / or the temperature error change rate are input into the pre-constructed fuzzy controller, further comprising: collecting a target temperature in the running of the target electronic device, and calculating a corresponding target temperature error based on the target temperature and a target expected temperature of the target electronic device, and calculating a target temperature error change rate of the target electronic device according to the target temperature error; determining a first fuzzy set corresponding to the target temperature error and a second fuzzy set corresponding to the target temperature error change rate based on the target temperature error and the target temperature error change rate; determining a third fuzzy value of the target heat dissipation object based on the first fuzzy set and the second fuzzy set, and using a rule base in the fuzzy controller; and constructing a fuzzy model in the fuzzy controller based on the first fuzzy set, the second fuzzy set, and the third fuzzy value.

[0094] It can be understood that, in the embodiments of the present application, the first fuzzy set and the second fuzzy set can but are not limited to include negative large, negative medium, negative small, zero, positive small, positive medium, positive large, etc., and the present application does not make specific limitations; the third fuzzy value can but is not limited to be negative large, negative medium, negative small, zero, positive small, positive medium, positive large, etc., and the present application does not make specific limitations. For example, negative large can be used to describe the direction (actual > expected) and amplitude (large gap) of the temperature error.

[0095] For example, the target temperature error , the target temperature error change rate , the target speed adjustment amount of the target heat dissipation object of the target server, such as the fan , wherein, represents negative large; represents negative medium; represents negative small; represents zero; represents positive small; represents positive medium; represents positive large.

[0096] In addition, the rule base in the fuzzy controller of the embodiments of the present application shows how the system dynamically adjusts the fan speed according to the temperature error and the temperature error change rate. These rules embody the core logic of fuzzy control: by comprehensively evaluating the error and its trend, dynamically adjusting the control strategy, and achieving smooth and accurate temperature regulation.

[0097] In some embodiments, the embodiments of the present application can collect a target temperature in the running of the target electronic device, and determine a corresponding target expected temperature, to calculate a corresponding target temperature error and a target temperature error change rate, and determine a third fuzzy value of the target heat dissipation object according to a first fuzzy set corresponding to the target temperature error, a second fuzzy set corresponding to the target temperature error change rate, and a corresponding rule base, thereby constructing a fuzzy model.

[0098] Before inputting the temperature error and / or temperature error change rate into the pre-constructed fuzzy controller, the embodiment of the application can calculate the target temperature error and change rate according to the target temperature and target expected temperature in the running of the target electronic device, and then determine the corresponding fuzzy set, and then determine the fan speed fuzzy value of the heat dissipation object by using the rule base, and then construct the fuzzy model of the fuzzy controller, and through dynamic target temperature adaptation, fuzzy set fine division and rule base self-optimization, a fuzzy model with strong adaptability, high robustness and low calculation overhead is constructed.

[0099] Optionally, in an embodiment of the application, before inputting the temperature error and / or temperature error change rate into the pre-constructed fuzzy controller, it further includes: constructing the membership function in the fuzzy controller based on the first fuzzy set and the second fuzzy set; calculating the first membership value of the target temperature error and the second membership value of the target temperature error change rate based on the membership function; analyzing the third fuzzy value based on the first membership value and the second membership value to determine the target adjustment amount of the target heat dissipation object; and constructing the defuzzification model in the fuzzy controller based on the membership function, the first membership value, the second membership value, the third fuzzy value and the target adjustment amount.

[0100] It can be understood that the embodiment of the application can construct the corresponding membership function by using the Gaussian membership function, and the expression can be but is not limited to:

[0101]

[0102] Among them, represents an input variable, which can be a temperature error or an error change rate; represents an input variable belongs to different fuzzy sets.

[0103] Further, the embodiment of the application can calculate the membership values of the target temperature error and the target temperature error change rate by using the membership function, and then realize the analysis of the fan speed fuzzy value and determine the corresponding target adjustment amount.

[0104] For example, the embodiment of the application is , represents an input variable belongs to a negative large membership degree, and the value range of the membership degree is between 0 and 1, and the value closer to 1 indicates that x is closer to the negative large; the value closer to 0 indicates that it is not close to the negative large.

[0105] Further, the embodiment of the application can use the gravity method for defuzzification to determine the target adjustment amount of the target heat dissipation object, and the calculation formula can be but is not limited to:

[0106]

[0107] ​​wherein, represents the total time of the fuzzy set elements participating in the defuzzification calculation; represents the total number of input variables (the value taken by the embodiment of the present application is 2), wherein, represents that the input variable is the temperature error, represents that the input variable is the temperature error rate of change; represents the index variable; represents the value corresponding to the first input variable at the first time; represents the membership degree of the corresponding fuzzy set, used to measure the degree of conformity with the fuzzy concept, and the value is between 0 and 1.

[0108] Further, the embodiment of the present application can construct a defuzzification model according to the membership function, the first membership value, the second membership value, the third fuzzy value and the target adjustment amount.

[0109] Before inputting the temperature error and / or the temperature error rate of change into the pre-constructed fuzzy controller, the embodiment of the present application constructs a membership function based on the first and second fuzzy sets, calculates the first and second membership values accordingly, analyzes the third fuzzy value to determine the target adjustment amount of the target heat dissipation object, and further constructs a defuzzification model, so that the adjustment amount of the heat dissipation object can be determined more accurately according to the temperature error and the rate of change, the rationality and accuracy of the fuzzy controller construction are improved, and the accuracy and effectiveness of the adjustment amount control are enhanced.

[0110] Optionally, in an embodiment of the present application, based on the first fuzzy set and the second fuzzy set, the third fuzzy value of the target heat dissipation object is determined by using the rule base in the fuzzy controller, including: based on the first fuzzy set and the second fuzzy set, determining the first fuzzy value of the target temperature error and the second fuzzy value of the target temperature error rate of change; comparing the first fuzzy value and the second fuzzy value with the rule base to obtain the third fuzzy value.

[0111] It can be understood that the rule base in the embodiment of the present application can include but is not limited to the following rules:

[0112] Rule 1: When the fuzzy value of the temperature error is positive large and the fuzzy value of the temperature error rate of change is negative large: it indicates that the current temperature is much lower than the target temperature, and the cooling trend is obvious. At this time, the embodiment of the present application can output a positive large control amount, that is, greatly increase the speed, flow rate, etc. of the heat dissipation object, which is not specifically limited by the present application, to accelerate cooling and prevent the temperature from continuously rising.

[0113] ​​Rule 2: when the fuzzy value of the temperature error is positive large and the fuzzy value of the temperature error rate of change is positive medium: it means that the current temperature is higher than the target temperature, but the temperature rising speed slows down. At this time, the embodiment of the present application can output a positive large control amount to maintain a high speed of the heat dissipation object, a flow rate, etc., to ensure that the temperature can be reduced to the preset stable interval in time.

[0114] Rule 3: when the fuzzy values of the temperature error and the temperature error rate of change are both around zero: it indicates that the current temperature is close to the target temperature and tends to be stable. At this time, the embodiment of the present application can output a zero control amount, i.e., keep the speed of the heat dissipation object, the flow rate, etc., unchanged, maintain the current heat dissipation state, and avoid temperature fluctuations caused by excessive adjustment.

[0115] Rule 4: when the fuzzy values of the temperature error and the temperature error rate of change are both negative large: it indicates that the temperature is far below the target temperature, and the temperature decreasing trend is significant. At this time, the embodiment of the present application can output a negative large control amount, i.e., significantly reduce the speed of the heat dissipation object, the flow rate, etc., to inhibit the temperature from further decreasing by reducing heat dissipation.

[0116] Other rules can also be used, which can be set by a person skilled in the art according to actual conditions, and the present application does not make specific limitations.

[0117] It can be understood that the embodiment of the present application can compare the first fuzzy value of the target temperature error and the second fuzzy value of the target temperature error rate of change with the rule base, and then obtain the third fuzzy value.

[0118] The embodiment of the present application determines the first and second fuzzy values of the target temperature error and the rate of change according to the first and second fuzzy sets, compares them with the rule base, and then obtains the third fuzzy value of the speed of the heat dissipation object of the target electronic device, the flow rate, etc., so as to accurately match the fuzzy information related to the temperature with the rule base, quickly and accurately deduce the fuzzy adjustment value of the speed of the heat dissipation object, the flow rate, etc., provide a reliable basis for subsequent accurate control of the speed of the heat dissipation object, the flow rate, etc., and improve the intelligence and accuracy of the electronic device heat dissipation control.

[0119] Optionally, in an embodiment of the present application, before the temperature of the electronic device is in the preset stable interval, it further includes: constructing a stability evaluation model of the electronic device based on the temperature error; determining a stability error value of the electronic device based on the stability evaluation model; and identifying the preset stable interval based on the stability error value.

[0120] It can be understood that the embodiment of the present application can use Lyapunov to prove stability, thereby constructing a corresponding stability evaluation model.

[0121] The process of Lyapunov stability proof can be: the embodiment of the present application proves the stability of the electronic device by the temperature error Construct an energy function The energy used to measure the system deviation from the steady state, its expression can be but not limited to:

[0122] ,

[0123] Where, because , so Always non-negative, if and only if (Temperature completely match the target temperature), when, .

[0124] Its derivative can be but not limited to:

[0125] ,

[0126] Where, , System parameters, reflecting the heat dissipation object adjustment (Such as speed adjustment, flow adjustment, etc., this application does not make specific restrictions) and temperature error On the error rate The influence of the intensity of the system itself is determined by the characteristics.

[0127] Because the fuzzy control rule satisfies:

[0128] ,

[0129] There is a constant , so that:

[0130] ,

[0131] Where, Indicates the exponential parameter, the value range is between Used to describe the heat dissipation object adjustment The rate of change of the error amplitude Make the adjustment more smooth.

[0132] So the system converges to steady state in finite time, steady state error , that is ,

[0133] Where, The stability error value is related to the fuzzy rule parameters, this parameter is obtained by experiment. When the rule base is fully optimized (such as fuzzy set division fine, high degree of membership function parameter matching, complete rule base and reasonable weight setting, etc., this application does not make specific restrictions), and the system hardware performance is good, the disturbance is small, the steady state error value Generally controllable in the range of 0.1~0.5. At this time, the embodiments of the present application can more accurately stabilize the temperature near the target temperature, meeting the higher precision requirements. If the rule base has certain optimization space (such as fuzzy set partitioning is relatively rough, part of the rule weight or membership function parameter setting is not accurate enough, etc., which is not specifically limited by the present application), or the system is slightly more disturbed, the steady-state error value Generally in the range of 0.5~1.5, in this case, the embodiments of the present application can still operate stably, but the precision of temperature control is slightly lower than the optimal case.

[0134] Before the embodiments of the present application make the temperature of the electronic device in a certain stable interval, a stability evaluation model is first constructed according to the temperature error, and then a stability error value is determined to identify a certain stable interval. The temperature error is used as the core to construct a quantitative stability evaluation model, which scientifically and accurately defines the temperature range required for stable operation of the electronic device, effectively improving the systematization and reliability of the temperature control of the electronic device.

[0135] In step S102, the temperature error and / or the temperature error change rate are input into the pre-constructed fuzzy controller to output the first adjustment amount of the heat dissipation object, and the heat dissipation object is adjusted according to the first adjustment amount to make the temperature of the electronic device in the preset stable interval.

[0136] It can be understood that in the case of the heat dissipation object being a fan, the first adjustment amount can be the adjustment amount of the fan speed, and in the case of the heat dissipation object being a cooling liquid, the first adjustment amount can be the adjustment amount of the cooling liquid flow rate. The specific settings can be made by those skilled in the art according to the actual situation, and the present application does not make specific limitations.

[0137] In some embodiments, the embodiments of the present application can use the pre-constructed fuzzy controller to determine the first adjustment amount of the heat dissipation object, and then adjust the heat dissipation object, so that the temperature of the electronic device is in a certain stable interval. The certain stable interval can be set by those skilled in the art according to the actual situation, and the present application does not make specific limitations.

[0138] For example, the embodiments of the present application can combine Figure 4 As shown in FIG. 1, in the case of the heat dissipation object being a fan, the first adjustment amount of the fan speed is generated, and the main content is as follows:

[0139] Step S401: Determine the expected temperature.

[0140] Step S402: Temperature sensor.

[0141] The embodiments of the present application can use the temperature sensor to collect the actual temperature of the server.

[0142] Step S403: Particle filtering denoising.

[0143] In the embodiments of the present application, the particle filter denoising can be used to obtain the final actual temperature. Figure 2

[0144] Step S404: temperature prediction.

[0145] In the embodiments of the present application, the temperature prediction can be performed by using the pre-constructed back propagation neural network to obtain the corresponding predicted temperature.

[0146] Step S405: calculation of temperature error and temperature error change rate.

[0147] Step S406: fuzzification.

[0148] In the embodiments of the present application, the temperature error and the temperature error change rate can be fuzzified to obtain the corresponding fuzzy value.

[0149] Step S407: rule base.

[0150] In the embodiments of the present application, the rule base can be used to determine the fuzzy value of the fan speed.

[0151] Step S408: defuzzification.

[0152] In the embodiments of the present application, the first adjustment amount of the fan speed can be determined by defuzzification.

[0153] Step S409: speed adjustment.

[0154] In the embodiments of the present application, the fan speed can be adjusted according to the first adjustment amount, so that the temperature of the electronic device is in a certain stable interval. The calculation formula of the current fan speed can be, but is not limited to:

[0155] ,

[0156] wherein, is the fan speed at the current moment, is the predicted temperature of the back propagation neural network, is the first adjustment amount, is the fan speed at the current moment. In step S103, a second adjustment amount of the heat dissipation object is generated according to the temperature and the power consumption of the electronic device, and the heat dissipation object is adjusted according to the second adjustment amount, so that the power consumption of the electronic device meets the preset optimization condition.

[0157] In step S103, a second adjustment amount of the heat dissipation object is generated according to the temperature and the power consumption of the electronic device, and the heat dissipation object is adjusted according to the second adjustment amount, so that the power consumption of the electronic device meets the preset optimization condition.

[0158] ​​It can be understood that the embodiment of the present application can reduce the heat dissipation power consumption of the electronic device through the hierarchical heat dissipation strategy and the dynamic sleep mechanism, and meet certain optimization conditions, wherein the certain optimization conditions can be set by the person skilled in the art according to the actual situation, and the present application does not make specific limitations.

[0159] In some embodiments, the embodiment of the present application can generate a second adjustment amount of the heat dissipation object according to the temperature and the heat dissipation power consumption of the electronic device, and then adjust the heat dissipation object, so that the heat dissipation power consumption of the electronic device meets certain optimization conditions.

[0160] Optionally, in an embodiment of the present application, generating a second adjustment amount of the heat dissipation object according to the temperature and the power consumption of the electronic device comprises: determining a preset stable interval corresponding to the temperature based on the temperature; and generating the second adjustment amount based on the preset stable interval and the power consumption.

[0161] It can be understood that the embodiment of the present application adopts a hierarchical heat dissipation strategy, so the embodiment of the present application can first determine a certain stable interval corresponding to different temperatures, and then generate a corresponding second adjustment amount.

[0162] In the embodiment of the present application, the certain stable interval can be divided into three levels, such as a first stable interval, temperature ≤ 25℃; a second stable interval, 25℃ < temperature ≤ 35℃; and a third stable interval, temperature > 35℃, which can be set by the person skilled in the art according to the actual situation, and the present application does not make specific limitations.

[0163] The embodiment of the present application determines a certain stable interval corresponding to the temperature of the electronic device, and then generates a second adjustment amount of the heat dissipation object in combination with the power consumption, so that the generated adjustment amount is more suitable for the actual operation requirements of the electronic device, avoids the inaccuracy caused by single factor adjustment, effectively improves the rationality and effectiveness of the heat dissipation control of the electronic device, and helps to ensure the stable operation of the electronic device and improve its performance and reliability.

[0164] Optionally, in an embodiment of the present application, generating a second adjustment amount based on a preset stable interval and power consumption comprises: determining the second adjustment amount as a first target value when the preset stable interval is a first stable interval; determining the second adjustment amount as a second target value when the preset stable interval is a second stable interval; determining the second adjustment amount as a third target value when the preset stable interval is a third stable interval; and determining the second adjustment amount as a fourth target value when the duration of the temperature in the corresponding preset stable interval is greater than a preset time.

[0165] In some embodiments, the embodiment of the present application can determine the second adjustment amount as a first target value when the certain stable interval is a first stable interval, such as maintaining the fan speed at 30%, and the present application does not make specific limitations.

[0166] In some embodiments, the embodiments of the present application can determine the second adjustment amount as the second target value when the certain stable interval is the second stable interval, such as maintaining the fan speed at 30%-70%, without specific limitation of the present application.

[0167] In some embodiments, the embodiments of the present application can determine the second adjustment amount as the third target value when the certain stable interval is the third stable interval, such as maintaining the fan speed at 70%-100%, full-speed heat dissipation, without specific limitation of the present application.

[0168] In addition, the embodiments of the present application adopt a dynamic hibernation mechanism, that is, when the temperature is stable in a certain stable interval for more than a certain time, such as 10 minutes, the fan hibernation mode is started, the speed is reduced by 20%, and the fourth target value is determined as the second adjustment amount. The certain time can be set by those skilled in the art according to the actual situation, without specific limitation of the present application.

[0169] The embodiments of the present application can set the corresponding second adjustment amount according to different certain stable intervals, and determine the fourth target value as the second adjustment amount when the temperature is in the corresponding interval for more than a preset time, which can provide more accurate and more actual demand- matching heat dissipation object adjustment amount according to the specific stable state of the electronic device temperature, better adapt to the subtle changes and stable trend of the electronic device temperature in the long-time running process, further optimize the heat dissipation effect, improve the stability and reliability of the electronic device running, make the heat dissipation object run at the appropriate speed as much as possible under the premise of meeting the heat dissipation demand of the electronic device, reduce unnecessary energy consumption, improve energy utilization efficiency, and reduce operating cost.

[0170] The working principle of the heat dissipation control method proposed by the embodiments of the present application will be introduced below in combination with a specific embodiment.

[0171] Among them, Figure 5 The flow chart of the working principle of the heat dissipation control method provided by an embodiment of the present application.

[0172] The method includes three stages.

[0173] Stage one: start-up preparation stage.

[0174] Step S501: power on the server.

[0175] Among them, the embodiments of the present application can connect the server to the power supply, start the hardware, load the system, and is the starting point of the flow.

[0176] Step S502: initialize parameters.

[0177] Wherein, the parameter initialization of the embodiment of the application is the basis of running, such as temperature sensor calibration value, initial weight of fuzzy control rule, particle number / distribution of particle filter, etc. The application does not make specific restrictions, and the subsequent calculation has a benchmark.

[0178] Step S503: particle filter initialization.

[0179] Wherein, the embodiment of the application can feed the particle filter algorithm with initial data, such as particle distribution near the initial temperature of the start-up, and the application does not make specific restrictions, so that it can start processing temperature data.

[0180] Step S504: back propagation neural network training.

[0181] Wherein, the embodiment of the application can use historical temperature data or simulation data of the same type of server as training temperature, or the server cluster of the same type used in a specific business scenario. The application does not make specific restrictions. Further, the embodiment of the application can collect data in the past year, a quarter, or a specific running period, and the application does not make specific restrictions.

[0182] Further, the embodiment of the application processes the collected data for data cleaning, standardization / normalization, etc., to obtain processed data, thereby determining training data, and training the back propagation neural network to learn to predict future trends according to past temperature changes, thereby obtaining the trained back propagation neural network, which is directly called in the subsequent.

[0183] Phase two: real-time closed-loop regulation phase.

[0184] Step S505: temperature sensor data collection.

[0185] Wherein, the embodiment of the application can use the temperature sensor to periodically read the internal temperature of the server (such as the temperature of the CPU, cabinet area, etc., which the application does not make specific restrictions), which is the sensing input of the regulation.

[0186] Step S506: particle filter denoising.

[0187] Wherein, the embodiment of the application can use particle filter to process the collected actual temperature, filter sensor errors, electromagnetic interference, etc. noise, to obtain the final actual temperature, so that the temperature value is more accurate.

[0188] Step S507: back propagation neural network prediction.

[0189] Wherein, the embodiment of the application can input the final actual temperature into the pre-constructed back propagation neural network, i.e. the trained back propagation neural network, for temperature prediction, and then obtain the predicted temperature and the corresponding temperature change trend in the future few seconds / minutes (e.g. the temperature in the future 10 seconds may rise by 2℃, which is not specifically limited by the application).

[0190] Step S508: Calculate the temperature error and the temperature error change rate.

[0191] Wherein, the embodiment of the application can calculate the temperature error by comparing the expected temperature and the final actual temperature, and then calculate the corresponding temperature error change rate in combination with the predicted temperature, to provide a decision basis for the fuzzy controller.

[0192] Step S509: The rule base in the fuzzy controller performs reasoning.

[0193] Wherein, the embodiment of the application can substitute the temperature error and the temperature error change rate into the rule base in the fuzzy controller, e.g. if the temperature error is large and the temperature error change rate is large, the fan speed is greatly increased, which is not specifically limited by the application, and then the fuzzy value of the fan speed is reasoned out.

[0194] Step S510: Obtain the first adjustment amount by defuzzification.

[0195] Wherein, the embodiment of the application can convert the third fuzzy value corresponding to the fan speed into a specific and executable adjustment amount, e.g. 30% increase in speed, etc., which is not specifically limited by the application, so as to enable the hardware to be executed.

[0196] Step S511: Control the fan to perform speed adjustment.

[0197] Wherein, the embodiment of the application issues a speed instruction to the fan, performs speed adjustment according to the first adjustment amount, directly adjusts the heat dissipation capacity, affects the actual temperature of the server, and then makes the temperature of the server in a certain stable interval.

[0198] Step S512: Determine whether the temperature of the server is in a certain stable interval.

[0199] Wherein, the embodiment of the application determines that the temperature is in a certain stable interval if the fluctuation is ≤0.5℃ for 5 consecutive minutes, and executes step S513, otherwise, executes step S505.

[0200] Step S513: Start the hierarchical heat dissipation strategy.

[0201] Wherein, the hierarchical heat dissipation strategy of the embodiment of the application divides a certain stable interval into three levels, such as a first stable interval, temperature≤25℃; a second stable interval, 25℃<temperature≤35℃; and a third stable interval, temperature>35℃, which can be set by a person skilled in the art according to actual conditions, and the application does not make specific limitations.

[0202] Step S514: starting the dynamic sleep mechanism.

[0203] Wherein, the embodiment of the application can reduce the fan speed when the temperature is stable, such as from full-power heat dissipation to low-power operation, or trigger the sleep strategy, reduce the power consumption of the heat dissipation system under the premise of ensuring that the temperature does not exceed the standard, so that the power consumption of the server meets certain optimization conditions.

[0204] In addition, it should be noted that the embodiment of the application can re-execute step S505 after stable regulation, continuous monitoring, and temperature fluctuation re-entering closed-loop regulation.

[0205] In summary, the embodiment of the application from server power-on, parameter initialization, to temperature sensing, particle filter denoising, back propagation neural network prediction, fuzzy controller decision, fan speed adjustment, server temperature stable cycle, finally optimizes the heat dissipation power of the server, realizes the closed-loop management of precise temperature control and energy saving and consumption reduction.

[0206] The embodiment of the application controls the steady-state error to be within ±0.3℃ by fusing the fuzzy controller and the particle filter, improves the accuracy by 40%, combines the prediction ability of the back propagation neural network, optimizes the response time to 5-7 seconds, improves the dynamic adjustment efficiency by nearly 50%, completely eliminates the chattering through the continuous adjustment mechanism of the fuzzy controller, reduces the fan speed fluctuation amplitude from ±15% to ±3%, reduces the heat dissipation power by 30-35% through the hierarchical heat dissipation strategy and the dynamic sleep mechanism, saves electricity cost, and when the sensor has a ±1℃ error, the control effect only decreases by 5% through the particle filter filtering and the prediction correction of the back propagation neural network.

[0207] Through the above description of the embodiments, those skilled in the art can clearly understand that the method according to the above embodiments can be realized by means of software and necessary general hardware platform, of course, it can also be realized by hardware, but in many cases, the former is a better embodiment.

[0208] According to the heat dissipation control method provided in the embodiments of the present application, the temperature error can be calculated according to the actual temperature and the expected temperature in the operation of the electronic device, the corresponding temperature error change rate can be calculated according to the temperature error, the first adjustment amount of the heat dissipation object can be output by using the pre-constructed fuzzy controller, and the adjustment can be performed, so that the temperature of the electronic device is in a certain stable interval, and then the second adjustment amount of the heat dissipation object can be generated according to the temperature and the power consumption of the electronic device, and the adjustment can be performed, so that the power consumption of the electronic device meets a certain optimization condition. Therefore, the problems that the PID controller faces in the nonlinear temperature system, such as response lag and large steady-state error, and the problems that the sliding mode control has, such as chattering effect, easy to cause fan mechanical loss, and then reduce the service life and reliability, can be solved, so as to achieve the technical effects of eliminating the chattering effect, improving the control smoothness, avoiding high-frequency switching, reducing the fan mechanical loss, reducing the state estimation error, improving the response speed, and shortening the response time.

[0209] The embodiments of the present application also provide a heat dissipation control device.

[0210] Figure 6 A block schematic diagram of the heat dissipation control device provided in the embodiments of the present application is shown.

[0211] As shown in Figure 6 The heat dissipation control device 10 is applied to an electronic device with a heat dissipation object, and the heat dissipation control device 10 comprises a first calculation module 100, a first adjustment module 200, and a second adjustment module 300.

[0212] The first calculation module 100 is configured to collect the actual temperature in the operation of the electronic device, calculate the corresponding temperature error based on the actual temperature and the expected temperature of the electronic device, and calculate the temperature error change rate of the electronic device according to the temperature error.

[0213] The first adjustment module 200 is configured to input the temperature error and / or the temperature error change rate into a pre-constructed fuzzy controller to output the first adjustment amount of the heat dissipation object, and adjust the heat dissipation object according to the first adjustment amount, so that the temperature of the electronic device is in a preset stable interval.

[0214] The second adjustment module 300 is configured to generate the second adjustment amount of the heat dissipation object according to the temperature and the power consumption of the electronic device, and adjust the heat dissipation object according to the second adjustment amount, so that the power consumption of the electronic device meets a preset optimization condition.

[0215] Optionally, in an embodiment of the present application, a second calculation module, a first determination module, a second determination module, and a first construction module are further included.

[0216] The second calculation module is configured to acquire a target temperature in the running of the target electronic device, and calculate a target temperature error and a target temperature error change rate of the target electronic device based on the target temperature and a target expected temperature of the target electronic device before inputting the temperature error and / or the temperature error change rate into the pre-constructed fuzzy controller.

[0217] The first determination module is configured to determine a first fuzzy set corresponding to the target temperature error and a second fuzzy set corresponding to the target temperature error change rate based on the target temperature error and the target temperature error change rate.

[0218] The second determination module is configured to determine a third fuzzy value of the target heat dissipation object by using a rule base in the fuzzy controller based on the first fuzzy set and the second fuzzy set.

[0219] The first construction module is configured to construct a fuzzy model in the fuzzy controller based on the first fuzzy set, the second fuzzy set and the third fuzzy value.

[0220] Optionally, in an embodiment of the present application, the method further comprises a second construction module, a third calculation module, a third determination module and a third construction module.

[0221] The second construction module is configured to construct a membership function in the fuzzy controller based on the first fuzzy set and the second fuzzy set before inputting the temperature error and / or the temperature error change rate into the pre-constructed fuzzy controller.

[0222] The third calculation module is configured to calculate a first membership value of the target temperature error and a second membership value of the target temperature error change rate based on the membership function.

[0223] The third determination module is configured to analyze the third fuzzy value based on the first membership value and the second membership value to determine a target adjustment amount of the target heat dissipation object.

[0224] The third construction module is configured to construct a defuzzification model in the fuzzy controller based on the membership function, the first membership value, the second membership value, the third fuzzy value and the target adjustment amount.

[0225] Optionally, in an embodiment of the present application, the second determination module comprises a first determination unit and a comparison unit.

[0226] The first determination unit is configured to determine a first fuzzy value of the target temperature error and a second fuzzy value of the target temperature error change rate based on the first fuzzy set and the second fuzzy set.

[0227] The comparison unit is configured to compare the first fuzzy value and the second fuzzy value with the rule base to obtain the third fuzzy value.

[0228] Optionally, in an embodiment of the present application, further comprising: a fourth constructing module, a fourth determining module and an identifying module.

[0229] The fourth constructing module is configured to construct a stability evaluation model of the electronic device based on the temperature error before the temperature of the electronic device is in the preset stable interval.

[0230] The fourth determining module is configured to determine a stability error value of the electronic device based on the stability evaluation model.

[0231] The identifying module is configured to identify the preset stable interval based on the stability error value.

[0232] Optionally, in an embodiment of the present application, the second adjusting module 300 comprises a second determining unit and a first generating unit.

[0233] The second determining unit is configured to determine the preset stable interval in which the temperature is based on the temperature.

[0234] The first generating unit is configured to generate the second adjusting amount based on the preset stable interval and the power consumption.

[0235] Optionally, in an embodiment of the present application, the first generating unit comprises a first determining subunit, a second determining subunit, a third determining subunit and a fourth determining subunit.

[0236] The first determining subunit is configured to determine that the second adjusting amount is a first target value in a case where the preset stable interval is a first stable interval.

[0237] The second determining subunit is configured to determine that the second adjusting amount is a second target value in a case where the preset stable interval is a second stable interval.

[0238] The third determining subunit is configured to determine that the second adjusting amount is a third target value in a case where the preset stable interval is a third stable interval.

[0239] The fourth determining subunit is configured to determine that the second adjusting amount is a fourth target value in a case where the duration in which the temperature is in the corresponding preset stable interval is greater than a preset time.

[0240] Optionally, in an embodiment of the present application, the first calculating module 100 comprises a second generating unit, a first calculating unit and a second calculating unit.

[0241] The second generating unit is configured to input the actual temperature into a pre-constructed back propagation neural network to obtain a predicted temperature of the electronic device.

[0242] The first calculating unit is configured to calculate the temperature error based on the actual temperature and the expected temperature, and / or the predicted temperature and the expected temperature.

[0243] The second calculation unit is configured to calculate a temperature error change rate based on the temperature error.

[0244] Optionally, in an embodiment of the present application, the method further comprises a first obtaining module, a fifth determining module and a fifth constructing module.

[0245] The first obtaining module is configured to obtain a training temperature corresponding to the back propagation neural network before the actual temperature is input to the pre-constructed back propagation neural network.

[0246] The fifth determining module is configured to determine a training function and a prediction function of the back propagation neural network based on the training temperature.

[0247] The fifth constructing module is configured to construct the back propagation neural network based on the training function and the prediction function.

[0248] Optionally, in an embodiment of the present application, the fifth determining module comprises a third determining unit, a fourth determining unit, a fifth determining unit, a sixth determining unit and a constructing unit.

[0249] The third determining unit is configured to determine a first weight matrix from an input layer to a hidden layer of the back propagation neural network based on the training temperature.

[0250] The fourth determining unit is configured to determine a first bias vector and a first activation function of the hidden layer based on the first weight matrix.

[0251] The fifth determining unit is configured to determine a second weight matrix from the hidden layer to an output layer of the back propagation neural network based on the first bias vector and the first activation function.

[0252] The sixth determining unit is configured to determine a second bias vector and a second activation function of the output layer based on the second weight matrix.

[0253] The constructing unit is configured to construct the prediction function based on the first weight matrix, the first bias vector, the first activation function, the second weight matrix, the second bias vector and the second activation function.

[0254] Optionally, in an embodiment of the present application, the method further comprises a second obtaining module, a third obtaining module and a sixth constructing module.

[0255] The second obtaining module is configured to obtain layer information of the back propagation neural network before the actual temperature is input to the pre-constructed back propagation neural network.

[0256] The third obtaining module is configured to obtain structure information of different layers of the back propagation neural network.

[0257] The sixth construction module is configured to construct the back propagation neural network based on the number-of-layers information and the structure information.

[0258] Optionally, in an embodiment of the present application, the first calculation module 100 comprises a seventh determination unit, a third calculation unit, an eighth determination unit, a ninth determination unit and a generation unit.

[0259] The seventh determination unit is configured to determine initial particle state values and initial particle weight values of different particles based on the actual temperature.

[0260] The third calculation unit is configured to calculate particle prediction state values of the different particles at different time instants based on the initial particle state values and the initial particle weight values.

[0261] The eighth determination unit is configured to determine particle prediction weight values of the different particles based on the particle prediction state values.

[0262] The ninth determination unit is configured to determine particles satisfying a preset particle condition based on the particle prediction weight values.

[0263] The generation unit is configured to obtain a final actual temperature based on the particles satisfying the preset particle condition.

[0264] Optionally, in an embodiment of the present application, the ninth determination unit comprises an acquisition subunit, a construction subunit, a generation subunit, a judgment subunit, a fifth determination subunit and a sixth determination subunit.

[0265] The acquisition subunit is configured to acquire particle expected weight values of the different particles.

[0266] The construction subunit is configured to construct effective evaluation models of the different particles based on the particle expected weight values.

[0267] The generation subunit is configured to input the particle prediction weight values into the effective evaluation models to obtain effective evaluation values of the different particles.

[0268] The judgment subunit is configured to judge whether the effective evaluation values are less than a preset threshold.

[0269] The fifth determination subunit is configured to determine that the different particles are invalid when the effective evaluation values are less than the preset threshold, and to re-generate the different particles to obtain newly generated particles, and to determine the particles satisfying the preset particle condition based on the newly generated particles.

[0270] The sixth determination subunit is configured to determine that the different particles are valid when the effective evaluation values are greater than or equal to the preset threshold, and to determine the particles satisfying the preset particle condition based on the valid particles.

[0271] The features of the embodiments of the heat dissipation control device can be referred to the related descriptions of the embodiments of the heat dissipation control method, which will not be repeated here.

[0272] The heat dissipation control device provided by the embodiments of the present application can calculate the temperature error according to the actual temperature and the expected temperature in the operation of the electronic device, calculate the corresponding temperature error change rate according to the temperature error, output the first adjustment amount of the heat dissipation object by using the pre-constructed fuzzy controller, and adjust the first adjustment amount, so that the temperature of the electronic device is in a certain stable interval, and then generate the second adjustment amount of the heat dissipation object according to the temperature and the power consumption of the electronic device, and adjust the second adjustment amount, so that the power consumption of the electronic device meets a certain optimization condition. Therefore, the problems of the PID controller in the nonlinear temperature system, such as response lag and large steady-state error, and the problems of the sliding mode control, such as chattering effect, easy to cause fan mechanical loss, and then reduce the service life and reliability, can be solved, so as to eliminate the chattering effect, improve the control smoothness, avoid high-frequency switching, reduce the fan mechanical loss, reduce the state estimation error, improve the response speed, and shorten the response time.

[0273] The embodiments of the present application also provide a controller, including a memory and a processor, the memory stores a computer program, and the processor is configured to run the computer program to execute the steps in any of the above-mentioned heat dissipation control method embodiments.

[0274] The embodiments of the present application also provide an electronic device including the above-mentioned controller, and the electronic device is used to implement the steps of any of the above-mentioned heat dissipation control methods.

[0275] The embodiments of the present application also provide a computer readable storage medium, which stores a computer program, and the computer program is configured to execute the steps in any of the above-mentioned heat dissipation control method embodiments when running.

[0276] In an example embodiment, the above-mentioned computer readable storage medium can include, but is not limited to, a U disk, a read-only memory (ROM), a random access memory (RAM), a mobile hard disk, a magnetic disk or an optical disk, and various media that can store computer programs.

[0277] The embodiments of the present application also provide a computer program product, which includes a computer program, and the computer program is executed by a processor to implement the steps in any of the above-mentioned heat dissipation control method embodiments.

[0278] The embodiment of the present application further provides another computer program product, comprising a nonvolatile computer readable storage medium, the nonvolatile computer readable storage medium stores a computer program, the computer program is executed by a processor to implement the steps in any of the above heat dissipation control method embodiments.

[0279] Those skilled in the art will further appreciate that the functions of the examples described herein-based units and algorithm steps can be implemented using electronic hardware, computer software, or any combination thereof. To clearly illustrate the interchangeability of hardware and software, various examples have been described herein in terms of their general functionality. Whether such functionality is implemented as hardware or software depends on the particular application and design constraints imposed on the overall system. Skilled artisans can implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present application.

[0280] The above has carried out the detailed introduction to the heat dissipation control method provided by the present application. The principle and implementation mode of the present application are described by applying specific examples herein, and the above embodiment description is only for helping to understand the method of the present application and its core idea. It should be pointed out that, for the ordinary skilled in the art, some improvements and modifications can be made to the present application without departing from the principle of the present application, and these improvements and modifications also fall within the protection scope of the claims of the present application.

Claims

1. A heat dissipation control method applied to an electronic device having a heat dissipation object, characterized by, The method comprises the following steps: collecting an actual temperature of the electronic device in operation, and calculating a corresponding temperature error based on the actual temperature and a desired temperature of the electronic device, and calculating a temperature error change rate of the electronic device according to the temperature error; inputting the temperature error and / or the temperature error change rate into a pre-constructed fuzzy controller to output a first adjustment amount of the heat dissipation object, and adjusting the heat dissipation object according to the first adjustment amount to make the temperature of the electronic device in a preset stable interval, wherein the input fuzzy set of the fuzzy controller is a first fuzzy set corresponding to the temperature error and a second fuzzy set corresponding to the temperature error change rate, and the output defuzzification value of the fuzzy controller is a third fuzzy value of the heat dissipation object; generating a second adjustment amount of the heat dissipation object according to the temperature and power consumption of the electronic device, and adjusting the heat dissipation object according to the second adjustment amount to make the power consumption of the electronic device meet a preset optimization condition; wherein, before making the temperature of the electronic device in a preset stable interval, further comprising: constructing a stability evaluation model of the electronic device based on the temperature error; determining a stability error value of the electronic device based on the stability evaluation model; identifying the preset stable interval based on the stability error value; the second adjustment amount of the heat dissipation object according to the temperature and power consumption of the electronic device, comprising: determining a preset stable interval where the temperature is based on the temperature; generating the second adjustment amount based on the preset stable interval and the power consumption; the second adjustment amount is generated based on the preset stable interval and the power consumption, comprising: in the case that the preset stable interval is a first stable interval, determining that the second adjustment amount is a first target value; in the case that the preset stable interval is a second stable interval, determining that the second adjustment amount is a second target value; in the case that the preset stable interval is a third stable interval, determining that the second adjustment amount is a third target value; in the case that the duration of the temperature in the corresponding preset stable interval is greater than a preset time, determining that the second adjustment amount is a fourth target value; the actual temperature of the electronic device in operation is collected, comprising: determining initial particle state values and initial particle weight values of different particles based on the actual temperature; calculating particle prediction state values of the different particles at different times based on the initial particle state values and the initial particle weight values; determining particle prediction weight values of the different particles based on the particle prediction state values; determining particles satisfying a preset particle condition based on the particle prediction weight values; obtaining a final actual temperature based on the particles satisfying the preset particle condition; the particles satisfying the preset particle condition are determined based on the particle prediction weight values, comprising: obtaining particle expected weight values of the different particles; constructing an effective evaluation model of the different particles based on the particle expected weight values; inputting the particle prediction weight values into the effective evaluation model to obtain effective evaluation values of the different particles; determining whether the effective evaluation value is less than a preset threshold value; if the effective evaluation value is less than the preset threshold value, determining that the different particle is invalid, and re-generating the different particle to obtain a new particle after re-generation, and determining the particle satisfying the preset particle condition based on the new particle; if the effective evaluation value is greater than or equal to the preset threshold value, determining that the different particle is valid, and determining the particle satisfying the preset particle condition based on the valid particle.

2. The method of claim 1, wherein, Before inputting the temperature error and / or the temperature error change rate into the pre-constructed fuzzy controller, further comprising: acquiring a target temperature of a target electronic device in operation, and calculating a corresponding target temperature error based on the target temperature and a target expected temperature of the target electronic device, and calculating a target temperature error change rate of the target electronic device according to the target temperature error; determining a first fuzzy set corresponding to the target temperature error and a second fuzzy set corresponding to the target temperature error change rate based on the target temperature error and the target temperature error change rate; determining a third fuzzy value of a target heat dissipation object based on the first fuzzy set and the second fuzzy set by using a rule base in the fuzzy controller; constructing a fuzzy model in the fuzzy controller based on the first fuzzy set, the second fuzzy set and the third fuzzy value.

3. The method of claim 2, wherein, Before inputting the temperature error and / or the temperature error change rate into the pre-constructed fuzzy controller, further comprising: constructing a membership function in the fuzzy controller based on the first fuzzy set and the second fuzzy set; calculating a first membership value of the target temperature error and a second membership value of the target temperature error change rate based on the membership function; analyzing the third fuzzy value based on the first membership value and the second membership value to determine a target adjustment amount of the target heat dissipation object; constructing a defuzzification model in the fuzzy controller based on the membership function, the first membership value, the second membership value, the third fuzzy value and the target adjustment amount.

4. The method of claim 2, wherein, The determining, based on the first fuzzy set and the second fuzzy set, of a third fuzzy value of a target heat dissipation object by using a rule base in the fuzzy controller, comprises: determining a first fuzzy value of the target temperature error and a second fuzzy value of the target temperature error change rate based on the first fuzzy set and the second fuzzy set; comparing the first fuzzy value and the second fuzzy value with the rule base to obtain the third fuzzy value.

5. The method of claim 1, wherein, The calculating, based on the actual temperature and an expected temperature of the electronic device, of a corresponding temperature error, and the calculating of a temperature error change rate of the electronic device according to the temperature error, comprises: inputting the actual temperature into a pre-constructed back propagation neural network to obtain a predicted temperature of the electronic device; calculating the temperature error based on the actual temperature and the expected temperature, and / or the predicted temperature and the expected temperature; calculating the temperature error change rate based on the temperature error.

6. The method of claim 5, wherein, Before inputting the actual temperature into the pre-constructed back propagation neural network, further comprising: obtaining a training temperature corresponding to the back propagation neural network; determining a training function and a prediction function of the back propagation neural network based on the training temperature; constructing the back propagation neural network based on the training function and the prediction function.

7. The method of claim 6, wherein, The determining of the prediction function of the back propagation neural network based on the training temperature comprises: determining a first weight matrix from an input layer to a hidden layer of the back propagation neural network based on the training temperature; determining a first bias vector and a first activation function of the hidden layer based on the first weight matrix; determining a second weight matrix from the hidden layer to an output layer of the back propagation neural network based on the first bias vector and the first activation function; determining a second bias vector and a second activation function of the output layer based on the second weight matrix; constructing the prediction function based on the first weight matrix, the first bias vector, the first activation function, the second weight matrix, the second bias vector and the second activation function.

8. The method of claim 5, wherein, Before inputting the actual temperature into the pre-constructed back propagation neural network, further comprising: obtaining layer number information of the back propagation neural network; obtaining structure information of different layers of the back propagation neural network; constructing the back propagation neural network based on the layer number information and the structure information.

9. A controller characterized by comprising: comprise: a memory, a processor, and a computer program stored in the memory and executable on the processor, the processor executing the program to implement the heat dissipation control method according to any one of claims 1-8.

10. An electronic device, comprising: The electronic device comprising the controller according to claim 9 is configured to implement the heat dissipation control method according to any one of claims 1-8.

Images