Control parameter determination method, performance adjustment method and related device of electronic equipment

Abstract

The invention discloses a control parameter determining method, a performance adjusting method and a related device of electronic equipment, and relates to the field of data processing, and the control parameter determining method of the electronic equipment comprises the following steps: obtaining a real-time processing parameter and a target processing parameter of a target task executed by the electronic equipment; acquiring a corresponding relationship between the real-time processing parameter and the target processing parameter; based on the corresponding relation, at least one first target parameter set is determined, and the first target parameter set comprises at least one first target parameter; on the basis of the first target parameter set, second target parameters are determined, and the second target parameters are related to execution resource scheduling of the target task.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and more specifically to a control parameter determination method, a performance adjustment method and related devices of an electronic device. Background Art

[0002] As the functions of electronic devices increase, the performance of electronic devices will affect the user experience when running applications. For example, when an electronic device processes display information, it will set a target refresh efficiency, but in the actual processing process, the real-time refresh efficiency will deviate from the target refresh efficiency. If it is not adjusted in time, it will affect the user experience or cause problems such as high power consumption of electronic devices. The current performance adjustment method will repeatedly adjust the parameters, which takes a long time and the adjustment effect is not good, resulting in low efficiency and poor accuracy in the performance adjustment of electronic devices. Summary of the invention

[0003] In view of this, this application provides the following technical solutions:

[0004] A method for determining a control parameter of an electronic device, comprising:

[0005] Obtaining real-time processing parameters and target processing parameters of a target task performed by the electronic device;

[0006] Acquiring a corresponding relationship between the real-time processing parameter and the target processing parameter;

[0007] Based on the corresponding relationship, determining at least one first target parameter group, wherein the first target parameter group includes at least one first target parameter;

[0008] Based on the first target parameter group, a second target parameter is determined, wherein the second target parameter is related to the execution resource scheduling of the target task.

[0009] Optionally, the corresponding relationship between the real-time processing parameter and the target processing parameter includes: a correlation coefficient between the real-time processing parameter and the target processing parameter;

[0010] The correlation coefficient is calculated based on the real-time processing parameter obtained by sampling at a target sampling number and the target processing parameter.

[0011] Optionally, the corresponding relationship between the real-time processing parameter and the target processing parameter includes: a correlation coefficient between the real-time processing parameter and the target processing parameter;

[0012] The method for forming the correlation coefficient includes: obtaining an initial correlation coefficient by calculating the real-time processing parameter obtained by sampling at a target sampling number and the target processing parameter; and obtaining the correlation coefficient by correcting the initial correlation coefficient based on a correction coefficient.

[0013] Optionally, when the real-time processing parameter is greater than the target processing parameter, the correction coefficient is obtained by at least calculating based on a deviation value between the real-time processing parameter and the target processing parameter.

[0014] Optionally, determining at least one first target parameter group based on the corresponding relationship includes:

[0015] Obtaining a dimension of a first parameter group of a controller, the controller being used to generate a second parameter based on the first parameter group; the first parameter group comprising at least one first parameter;

[0016] Determine the corresponding relationships between the first parameters within the candidate first parameter range of each dimension;

[0017] Based on the target screening conditions corresponding to the plurality of corresponding relationships corresponding to the first parameters, determine the first target parameter of each dimension from the candidate first parameter range in each dimension;

[0018] A combination of the first target parameters in various dimensions is taken as the first target parameter group.

[0019] Optionally, the determining corresponding coefficients corresponding to the first parameters in the candidate first parameter range of each dimension includes:

[0020] Based on the priority of each dimension, each first parameter within the candidate first parameter range of each dimension is traversed in turn to obtain the corresponding relationships between the first parameters in each dimension.

[0021] Optionally, based on the priority of each dimension, each first parameter within the candidate first parameter range of each dimension is traversed in turn to obtain corresponding relationships corresponding to the first parameters in each dimension; based on the target screening conditions corresponding to the corresponding relationships corresponding to the first parameters, the first target parameter of each dimension is determined from the candidate first parameter range in each dimension, including:

[0022] Determine the target dimension of the current traversal, and keep the first parameters of other dimensions unchanged, wherein the first parameters of the traversed dimensions among the other dimensions are configured as the first target parameters determined under the dimension;

[0023] Traversing each first parameter under the target dimension, so that the controller generates a corresponding plurality of second parameters in sequence;

[0024] Based on the real-time processing parameters and target processing parameters of the target task corresponding to each of the second parameters, the corresponding relationships corresponding to each of the first parameters under the target dimension are obtained;

[0025] Determine the first target parameter of the target dimension based on the target screening conditions corresponding to the plurality of corresponding relationships corresponding to the first parameters under the target dimension;

[0026] According to the above steps, other dimensions are traversed in sequence until the first target parameters in all dimensions are determined.

[0027] Optionally, the controller comprises a PID controller.

[0028] A performance adjustment method, comprising:

[0029] Obtaining current real-time processing parameters of a target task executed by an electronic device;

[0030] In response to the real-time processing parameter satisfying an update condition, updating the second target parameter based on the first target parameter group; wherein the first target parameter group is determined based on a correspondence between the real-time processing parameter and the target processing parameter;

[0031] The scheduling of the execution resources of the target task is controlled according to the updated second target parameter.

[0032] A device for determining a control parameter of an electronic device, comprising:

[0033] A first acquisition unit, used to obtain real-time processing parameters and target processing parameters of a target task executed by the electronic device;

[0034] A second acquisition unit, used for acquiring a correspondence between the real-time processing parameter and the target processing parameter;

[0035] A first determining unit, configured to determine at least one first target parameter group based on the corresponding relationship, wherein the first target parameter group includes at least one first target parameter;

[0036] The second determination unit is used to determine a second target parameter based on the first target parameter group, wherein the second target parameter is related to the execution resource scheduling of the target task.

[0037] A performance adjustment device, comprising:

[0038] A third acquisition unit, used to obtain the current real-time processing parameters of the target task executed by the electronic device;

[0039] An updating unit, configured to update a second target parameter based on a first target parameter group in response to the real-time processing parameter satisfying an update condition; wherein the first target parameter group is determined based on a correspondence between the real-time processing parameter and the target processing parameter;

[0040] A control unit is used to control the scheduling of execution resources of the target task according to the updated second target parameter. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are merely embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0042] Figure 1 A schematic diagram of a flow chart of a method for determining control parameters of an electronic device provided in an embodiment of the present application;

[0043] Figure 2 A schematic diagram of a calculation process of a correlation coefficient provided in an embodiment of the present application;

[0044] Figure 3 A flowchart of a method for determining a target parameter group provided in an embodiment of the present application;

[0045] Figure 4 A schematic diagram of a performance adjustment method provided in an embodiment of the present application;

[0046] Figure 5 A schematic diagram of an Android system dynamic performance framework provided in an embodiment of the present application;

[0047] Figure 6 A flow chart of a method for determining PID parameters provided in an embodiment of the present application;

[0048] Figure 7 A schematic diagram of the structure of a control parameter determination device for an electronic device provided in an embodiment of the present application;

[0049] Figure 8 A schematic diagram of the structure of a performance adjustment device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0050] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0051] The terms "first" and "second" in this application are used to distinguish different objects, rather than to describe a specific order. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but may include steps or units that are not listed.

[0052] In the process of electronic devices performing related tasks, in order to ensure the normal operation of the tasks and the performance of the electronic devices, it is usually necessary for the electronic devices to maintain within the corresponding energy consumption ratio range when performing the related tasks. Therefore, the control parameter determination method of the electronic device provided in the embodiment of the present application can be executed to determine the corresponding parameters to adjust the scheduling resources corresponding to the target task, thereby improving the performance of the electronic device when processing the target task. Among them, the electronic devices in the embodiments of the present application may include mobile phones, computers, virtual reality devices, vehicle-mounted terminals or wearable devices, etc.

[0053] See also Figure 1 , is a flow chart of a method for determining control parameters of an electronic device provided in an embodiment of the present application, the method may include the following steps:

[0054] S101. Obtain real-time processing parameters and target processing parameters of a target task executed by an electronic device.

[0055] S102: Obtain the corresponding relationship between the real-time processing parameters and the target processing parameters.

[0056] S103: Determine at least one first target parameter group based on the corresponding relationship.

[0057] S104. Determine a second target parameter based on the first target parameter group.

[0058] In step S101, the electronic device may be a device with a data processing function, and the target task executed by the electronic device may be one of the tasks currently running in the electronic device, or a task to be monitored determined based on the current running load of the electronic device, or a task determined according to the priority of each task in the electronic device. For example, in order to improve the user experience, the priority of the task corresponding to the user interface of the electronic device is usually higher, such as the target task may be a task of displaying information rendering, etc.

[0059] In order to enable the electronic device to have a better energy consumption ratio when executing the target task, it is necessary to monitor the processing parameters of the electronic device when executing the target task to obtain the real-time processing parameters. Further, in order to enable the target task to have a better processing effect during the execution process, the corresponding target processing parameters will be pre-set, that is, when the target task is executed with the target processing parameters, the task processing effect is better, and the performance of the electronic device is also in a better state. Depending on the type of the target task, the corresponding processing parameters may also be different. For example, when the target task is an image rendering task, the processing parameters may include the time to render a frame of image, etc. If the target task is a data calculation task, the processing parameters may be computing efficiency, etc. Correspondingly, if the target task is an image rendering task, the corresponding real-time processing parameters may be the actual time to complete the rendering of a frame of image, and the target processing parameters may be the expected time to complete the rendering of a frame of image. In this way, the processing status of the target task can be obtained according to the real-time processing parameters of the target task, which is convenient for the subsequent adjustment of the relevant information of the target processing task.

[0060] After obtaining the real-time processing parameters of the target task and the corresponding target processing parameters, the corresponding relationship between the real-time processing parameters and the target processing parameters can be determined in step S102, so as to determine whether the current real-time processing parameters meet the processing requirements of the target task, and the energy consumption status of the electronic device, etc. Specifically, the corresponding relationship between the real-time processing parameters and the target processing parameters can include the deviation relationship between the real-time processing parameters and the target processing parameters, and can also include the size relationship between the real-time processing parameters and the target processing parameters, etc., and further, can also include the relationship between the real-time processing parameters and the target processing parameters and the current resource occupancy, the energy consumption of the electronic device, etc.

[0061] When determining the processing parameters in the embodiments of the present application, the processing parameters can be determined based on the task characteristics of the target task of interest. For example, when the current electronic device is mainly running a video display application, the target task is a data display task, and its corresponding processing parameters can be parameters that affect the display effect, such as image rendering parameters, image encoding and decoding parameters, etc. Correspondingly, the various operating parameters of the target task can also be parsed, and the operating parameters that can affect the operating effect of the target task can be determined as processing parameters that need to be monitored. For example, if the target task is a data reading task, the corresponding processing parameters can be data reading efficiency parameters or data transmission efficiency parameters, etc.

[0062] Correspondingly, when determining the processing parameters in the embodiment of the present application, it can also be determined according to the current application scenario of the electronic device. If the electronic device is in a high-power operation state, the operation parameters corresponding to the processing tasks that currently affect the power consumption of the electronic device can be determined as the processing parameters. Therefore, the electronic device can be ensured to be in a normal operating state based on the monitoring of the real-time processing parameters. Specifically, it can be determined whether the current real-time processing parameters need to be adjusted by monitoring the corresponding relationship between the real-time processing parameters and the target processing parameters, thereby ensuring the operating performance of the electronic device.

[0063] Through step S103, at least one first target parameter group can be determined based on the corresponding relationship, and the first target parameter value includes at least one first target parameter. In the embodiment of the present application, multiple parameter groups can be determined according to the real-time processing parameters and the target processing parameters, and each parameter group includes multiple parameters of different dimensions. These parameter groups are used to make the controller output corresponding control parameters, and the resources corresponding to the target task are adjusted by the control parameters. Thereby, the processing parameters corresponding to the target task can be changed to meet the conditions corresponding to the target processing parameters.

[0064] Specifically, based on the type or working principle of the controller, the corresponding parameter dimensions in the parameter group may be different. Controllers can be divided into different types of controllers, such as feedback controllers, feedforward controllers, sequential controllers, neural network controllers, and program controllers. Among them, the feedback controller uses the output information of the system to adjust the control input so that the actual output of the system is close to the expected set value, that is, the output feedback control signal is calculated by internal parameters. The parameter dimensions in the parameter groups of different types of controllers are different. Correspondingly, even if the controllers of the same type have different corresponding control logics, the corresponding parameter dimensions in the parameter groups may also be different. For example, feedback controllers may include proportional controllers, integral controllers, differential controllers, proportional integral controllers (PI controllers), proportional differential controllers (PD controllers), and proportional integral differential controllers (PID controllers). For example, the parameter dimensions in the parameter group corresponding to the PI controller may include proportional dimensions and integral dimensions; the parameter dimensions in the parameter group corresponding to the PD controller may include proportional dimensions and differential dimensions.

[0065] Taking the proportional integral derivative (PID) controller as an example, the parameters in the parameter group may be a proportional parameter, an integral parameter, and a differential parameter. The PID controller determines a target parameter group with different proportional, integral, and differential parameters according to the corresponding relationship between the real-time processing parameters and the target processing parameters, and determines the target parameter group that best matches the current demand from these target parameter groups as the first target parameter group. Then, according to each first target parameter in the first target parameter group, the controller outputs a second target parameter, and the second target parameter is used as a control parameter for controlling the target task to adjust the resources. That is, the second target parameter is related to the execution resource scheduling of the target task. For example, the execution resources of the target task may include the processing thread corresponding to the target task, and may also include the computing resources of the target task, etc. The second target parameter may be a related parameter for adjusting the working parameters of the processing thread. Correspondingly, the second target parameter may also be a parameter for controlling resource calls, etc. Correspondingly, the scheduling resources corresponding to the target task may include the computing thread resources, access resources, and the read amount of cache resources corresponding to the target task, etc.

[0066] In addition to the control and adjustment of parameters of commonly used parameter dimensions in each controller, control can also be performed based on parameters of other dimensions according to the working principle of the controller and the current control requirements. For example, in a PID controller, the parameter dimensions in the parameter group include not only the proportional, integral and differential dimensions, but also the limiting parameter dimension. In order to prevent the controller output from being too large or too small, the PID controller can also be provided with an output limiting parameter dimension, which limits the maximum and minimum values ​​of the controller output. A sampling time dimension can also be provided to set the sampling time through this dimension, thereby determining the frequency at which the controller updates the control output, so that the corresponding relationship between the real-time processing parameters and the target processing parameters can be more accurately determined through the parameters of the sampling time dimension, thereby facilitating the adjustment of the parameters.

[0067] Correspondingly, in the embodiment of the present application, the controller can also be a program controller, such as a thread controller for managing and controlling processing threads, wherein the processing threads are used to execute the corresponding processing flow when the target task is running. If the current real-time processing parameters characterize that the processing efficiency of the target task is low, the processing threads for the target task can be increased through the thread controller. However, if a large number of processing threads are added, it will affect the thread calls of other tasks. Therefore, in the embodiment of the present application, the parameter range of the thread scheduling parameters of the thread controller can be determined based on the correspondence between the real-time processing parameters and the target processing parameters, such as the range of the number of processing threads that can be added, etc. Then the thread controller can be controlled based on the number of processing threads within the parameter range to determine the target number of processing threads to be added, so as to adjust the processing efficiency of the target task.

[0068] For another example, the controller may also include a controller based on an artificial intelligence model, which can adjust the real-time processing parameters. However, if the controller is used to directly adjust the real-time processing parameters of the target task, the adjustment process of the real-time processing parameters will be cumbersome and inefficient due to the complex structure of the artificial intelligence model and the model weight parameters. Therefore, based on the correspondence between the real-time processing parameters and the target processing parameters, the weight parameters of the artificial intelligence model corresponding to the controller can be determined, so that the controller can adjust the information related to the resource scheduling of the target task according to the weight parameters to achieve the purpose of adjusting the real-time processing parameters, thereby improving the working efficiency of the controller and the adjustment accuracy of the corresponding processing parameters of the target task.

[0069] In an embodiment of the present application, an analysis can be performed based on the correspondence between the real-time processing parameters and the target processing parameters to obtain an optimal first target parameter group, which is used by a controller to determine a second target parameter. The controller is used to adjust the control variable. The second target parameter is related to the execution resource scheduling of the target task. In this way, the availability of the target parameters included in each target parameter group can be evaluated in real time based on the correspondence between the real-time processing parameters and the target processing parameters, thereby determining the optimal first target parameter group without manually evaluating each target parameter group. At the same time, the correspondence between the real-time processing parameters and the target processing parameters is used as a judgment condition for determining the first target parameter group, without manually determining the evaluation criteria, so that the parameter determination process is more accurate and can better meet actual application needs.

[0070] The following describes a method for determining control parameters of an electronic device corresponding to an embodiment of the present application in conjunction with a specific application scenario.

[0071] In one implementation of the present application, the correspondence between the real-time processing parameter and the target processing parameter includes: a correlation coefficient between the real-time processing parameter and the target processing parameter. The correlation coefficient is calculated based on the real-time processing parameter and the target processing parameter obtained by sampling at the target sampling number.

[0072] In order to accurately determine the correspondence between the real-time processing parameters and the target processing parameters, the correlation coefficient between the two can be determined through several samplings, that is, the real-time processing parameters and the target processing parameters are obtained through each sampling, and then the correlation coefficient between the two is determined through the processing mode of average calculation.

[0073] For example, the target sampling number is represented by n, which means that n samplings are performed, and the real-time processing parameters obtained by the i-th sampling are represented by x. i Indicates that the corresponding target processing parameter is y iIndicates that a correlation coefficient is calculated as follows, expressed as Corr:

[0074] (1)

[0075] It can be seen that the correlation coefficient can indicate the closeness between the real-time processing parameters and the target processing parameters. When the real-time processing parameters are close to the target processing parameters, the correlation coefficient is closer to 1, which can meet the processing requirements of the target tasks of the electronic device, that is, it can meet the performance of the electronic device to execute the target tasks, and there will be no excess performance, thereby saving the processing power consumption of the electronic device.

[0076] Correspondingly, the correlation coefficient between the real-time processing parameter and the target processing parameter can be obtained by using the corresponding correlation coefficient calculation method according to the current application scenario and the parameter characteristics of the real-time processing parameter and the target processing parameter. For example, the real-time processing parameter and the target processing parameter are two linearly related quantities, and the Pearson correlation coefficient calculation method can be used to calculate the relationship between the real-time processing parameter and the target processing parameter. As shown in formula (2):

[0077] (2)

[0078] In the above formula (2), X can represent the real-time processing parameter, Y can represent the target processing parameter, r(X,Y) represents the correlation coefficient; Cov(X,Y) is the covariance of X and Y, is the variance of X, is the variance of Y. The larger the absolute value of the correlation coefficient r (the closer it is to 1), the higher the degree of linear correlation between the two quantities; the smaller the absolute value of the correlation coefficient r, the lower the degree of linear correlation between the two quantities.

[0079] For another example, if the real-time processing parameter and the target processing parameter are nonlinear or do not satisfy the normal distribution, the Spearman rank correlation coefficient can also be used to determine the correlation coefficient between the two. Correspondingly, the Kendall rank correlation coefficient can also be selected according to the characteristics of the parameters to calculate the correlation coefficient between the two. The embodiment of the present application does not limit the method for calculating the correlation coefficient between the real-time processing parameter and the target processing parameter, and only needs to select the corresponding calculation method according to the data characteristics and application scenarios.

[0080] Correspondingly, in addition to being characterized by the above-mentioned correlation coefficient, the correspondence between the real-time processing parameters and the target processing parameters can also be characterized by the absolute value of the deviation between the two, and the correspondence between the two parameters can also be represented by covariance, where the covariance can represent the degree of change of the two parameters. The correspondence between the parameters of the chain can also be represented by regression calculation values, etc. Specifically, the representation method of the correspondence between the two can be determined according to the data characteristics of the processing parameters and the current scene analysis requirements, and this application does not limit this.

[0081] Taking the target task as an image rendering task, the processing parameter represents the time to render a frame of image. Assuming that the current real-time rendering time of a frame of image is 0.019 seconds, and the expected rendering time of a frame of image is 0.018 seconds, although the two are very close, since the real-time rendering time of a frame of image is greater than the expected rendering time of a frame of image, it may still cause display freezes and other problems. Therefore, in order to solve the above problem, in the process of calculating the correlation coefficient in the embodiment of the present application, the process of adjusting the initial correlation coefficient obtained by the above method by means of a correction coefficient is also included, so that the correlation coefficient obtained by the final calculation can better reflect the corresponding relationship between the current real-time processing parameters and the target processing parameters.

[0082] For another example, the target task may be a data reading task, and the processing parameter characterizes the data reading efficiency, which may specifically be the time to read the target length data. In an efficient data transmission scenario, if the data reading efficiency is low, it will affect the data transmission and subsequent data processing. At this time, in order to ensure the data reading and transmission effect, it is necessary to set the target efficiency parameter of the data reading. If the current real-time reading efficiency parameter is less than the target reading efficiency parameter, the reading effect will be reduced. If it is greater than the target reading efficiency parameter, it will occupy too much data reading resources and affect the operation of the electronic device. Therefore, in the embodiment of the present application, the real-time reading efficiency parameter needs to be close to the target reading efficiency parameter to ensure that the data reading effect and the energy consumption ratio of the electronic device are in a good state. Specifically, through the corresponding relationship between the real-time reading efficiency parameter and the target reading efficiency parameter, the control parameter that can be used to update the reading efficiency parameter is determined, so as to adjust the reading efficiency and meet the needs of the data reading task.

[0083] In one implementation of the embodiment of the present application, the correspondence between the real-time processing parameter and the target processing parameter includes: a correlation coefficient between the real-time processing parameter and the target processing parameter. A method for forming the correlation coefficient includes: obtaining an initial correlation coefficient by calculating the real-time processing parameter and the target processing parameter obtained by sampling at a target sampling number; and obtaining the correlation coefficient by correcting the initial correlation coefficient based on a correction coefficient.

[0084] In relevant application scenarios, the running process of the target task will also affect the overall state of the electronic device (such as affecting the power consumption of the electronic device, etc.). Therefore, it is necessary to correct the initial correlation coefficient of the real-time processing parameter and the target processing parameter based on the characteristics of these scenarios, that is, to characterize the trade-off between the performance of the target task and the power consumption of the electronic device by the correction coefficient, so that the correlation coefficient finally determined can accurately characterize the relationship between the real-time processing parameter and the target processing parameter in the current scenario. For example, when the real-time processing parameter is greater than the target processing parameter, even if the two are very close, they usually do not meet the actual application requirements. This is because when setting the target processing parameter, not only the processing effect of the target task is considered, but also the running effect of the electronic device that executes the target task is considered. When the real-time processing parameter is greater than the target processing parameter, it can affect the running effect of the electronic device and cannot achieve a good energy consumption ratio. Taking the target task as an image rendering task as an example, if the real-time rendering efficiency is greater than the target rendering efficiency, the corresponding rendering process will occupy too many processing resources of the electronic device, affecting the power consumption of the electronic device or the running of other tasks, such as causing the electronic device to consume power faster, or affecting the display effect of the electronic device display interface, etc., which affects the running effect and power consumption of the electronic device. Therefore, when the real-time processing parameter is greater than the target processing parameter, the correlation coefficient obtained by the above calculation needs to be adjusted by a corresponding correction coefficient to more accurately reflect the relationship between the real-time processing parameter and the target processing parameter.

[0085] Specifically, when the real-time processing parameter is greater than the target processing parameter, the correction coefficient is obtained by calculating at least the deviation value between the real-time processing parameter and the target processing parameter. For example, the real-time processing parameter is represented by x, the target processing parameter is represented by y, and the correction coefficient is represented by k, then:

[0086] k=((xy) / 2) 2 (3)

[0087] At this time, based on formula (1), the calculation method of the correlation coefficient can be changed to:

[0088] (4)

[0089] It should be noted that the above-mentioned calculation method of the correction coefficient k is an optional method provided by the embodiment of the present application. The correction coefficient k represents the trade-off between performance and power consumption when the target task is running. Therefore, the correction coefficient can be adjusted to meet different scenario requirements, such as meeting different requirements of performance priority, power consumption priority, or performance and power consumption balance. Therefore, the correction coefficient can also be set to a corresponding constant value according to actual application requirements; or the relevant weight parameters are set based on the value range corresponding to the difference between the real-time processing parameter and the target processing parameter. For example, let the constant be c, and the weight parameter be b when the difference is in the first numerical range. At this time, the correction coefficient k is the product of b and c; when the difference is in the second numerical range, the weight parameter is d, and at this time, the correction coefficient k is the product of d and c.

[0090] Further, the value range of the correction coefficient can also be set. If the value range includes four values, the first value, the second value, the third value and the fourth value are in descending order. Then, according to the current scene requirements, the corresponding value is obtained within the value range of the correction coefficient. For example, when the electronic device performs the target task, its operating state is good, and the current demand is to prioritize the operating performance, then the correction coefficient can select a smaller value, if the fourth value can be selected. If the current power of the electronic device is relatively low and it is necessary to ensure that the power consumption of the electronic device is normal, the correction coefficient can select a larger value, such as selecting the first value; for another example, it is necessary to ensure the balance between the performance and energy consumption of the electronic device, and the correction coefficient can select an intermediate value, such as the second value or the third value. Therefore, in the embodiment of the present application, different application requirements can be met by the correction coefficient, so that the corresponding relationship between the determined real-time processing parameters and the target processing parameters is more matched with the current application scenario, and further makes the subsequent processing more meet the needs of the current scenario.

[0091] See also Figure 2 , is a schematic diagram of a calculation process of a correlation coefficient provided in an embodiment of the present application, and the calculation process of the correlation coefficient may include the following steps:

[0092] S201: Acquire current real-time processing parameters and target processing parameters of a target task executed by an electronic device.

[0093] S202: Determine whether the real-time processing parameter is greater than the target processing parameter. If yes, execute step S203; if not, execute step S206.

[0094] S203: Calculate and obtain an initial correlation coefficient based on the real-time processing parameters and the target processing parameters obtained by sampling at the target sampling times.

[0095] S204: Calculate and obtain a correction coefficient based on the deviation value between the current real-time processing parameter and the target processing parameter.

[0096] S205. Correct the initial correlation coefficient based on the correction coefficient to obtain the correlation coefficient.

[0097] S206: Calculate and obtain a correlation coefficient based on the real-time processing parameters and the target processing parameters obtained by sampling at the target sampling number.

[0098] Specifically, in the process of calculating the correlation coefficient, the calculation method of the correction coefficient in step S204 can refer to the formula (3) in the above embodiment. The calculation process of the correlation coefficient in step S205 can refer to the calculation method of formula (4) in the above embodiment. The calculation process of the correlation coefficient in step S206 can refer to the calculation method of formula (1) in the above embodiment, which will not be described in detail here.

[0099] After obtaining the correlation coefficient that characterizes the corresponding relationship between the real-time processing parameters and the target processing parameters, at least one target parameter group can be determined according to the correlation coefficient. In order to accurately determine the parameters in each target parameter group so as to be used in the subsequent controller calculation process, it is necessary to calculate according to the parameter dimension of the controller.

[0100] In one implementation of the present application, a method for determining a target parameter group is also provided. Figure 3 , the method may include the following steps:

[0101] S301, obtaining the dimension of a first parameter group of a controller.

[0102] S302: Determine corresponding relationships between first parameters within a candidate first parameter range of each dimension.

[0103] S303: Based on the target screening conditions corresponding to the corresponding relationships corresponding to the first parameters, determine the first target parameters of each dimension from the candidate first parameter ranges in each dimension.

[0104] S304: Take the combination of the first target parameters of each dimension as the first target parameter group.

[0105] In this embodiment, the controller is used to generate a second parameter based on the first parameter group. The first parameter group includes at least one first parameter. The number of first parameters in the specific first parameter group is related to the dimension of the first parameter group. For example, the first parameter group of the controller includes three dimensions, and the corresponding first parameter group may include three first parameters. Specifically, the controller is used to calculate and obtain the second parameter according to each parameter in the first target parameter group, so that the real-time processing parameter can meet the matching condition corresponding to the target processing parameter based on the second parameter, thereby improving the processing efficiency of the target task. For example, the controller can be a proportional integral differential PID controller, and the target task can be a display rendering task of a GPU (graphics processor) of an electronic device. The PID controller can effectively ensure the load balancing of GPU rendering and minimize the total power consumption of the system under the premise of ensuring the quality of service. The PID controller can determine the proportional term, the integral term and the differential term based on the corresponding relationship between the real-time processing parameter of the controlled object (i.e., the target task) and the target processing parameter, so as to adjust the relevant resources or processing threads of the controlled object so that the real-time processing parameter corresponding to the controlled object gradually stabilizes within the range of the target processing parameter.

[0106] Among them, the target screening condition is to screen out the conditions that are most relevant to the real-time processing parameters and the target processing parameters characterized by the corresponding relationship. For example, the corresponding relationship is determined by the deviation value of the real-time processing parameters and the target processing parameters corresponding to each first parameter, and the target screening condition can be the condition with the smallest deviation value. Taking the corresponding relationship as the correlation coefficient as an example, the target screening condition can be the condition with the largest correlation coefficient. Therefore, in an embodiment of the present application, the target screening condition can be set according to the determination method of the corresponding relationship, so that the parameters obtained by screening based on the target screening condition are more accurate.

[0107] Different controller types or working principles have different corresponding parameter dimensions. Furthermore, the same type of controller may have different corresponding parameter dimensions in different application scenarios. The specific need is to determine the corresponding parameter dimensions of the controller according to the real-time application requirements and the characteristics of the target task. Taking the PID controller as an example, the dimensions of the corresponding first parameter group may be proportional, integral, and differential dimensions. Furthermore, corresponding to the processing process of the PID controller generating the second parameter, the dimensions of the first parameter group may be further refined into proportional, integral, and differential dimensions that can control the second parameter to rise (for example, control the second parameter to become larger), and proportional and differential dimensions that can control the second parameter to fall. It may also include a parameter dimension for determining how much data to take when calculating the second parameter value (the parameter dimension may also include corresponding proportional, integral, and differential dimensions). Among them, the parameter dimension for determining how much data to take when calculating the second parameter value defines the sampling window size for calculating the processor utilization of the electronic device. The specific value should be determined according to the specific requirements and response time of the system to ensure that the system can respond quickly to load changes while avoiding performance fluctuations caused by overreaction. Specifically, the dimension of the first parameter of the controller can be determined according to the working principle of the controller and the task requirements of the target task.

[0108] In the embodiment of the present application, the corresponding relationship can be specifically a parameter that can characterize the relationship between the real-time processing parameter and the target processing parameter, such as the various correlation coefficients introduced in the above embodiments. For example, the corresponding relationship can be characterized by the deviation value between the real-time processing parameter and the target processing parameter, the corresponding relationship can also be characterized by the correlation coefficient between the two, and the corresponding relationship can also be characterized by the similarity parameter between the two.

[0109] The method for calculating the corresponding relationship can be used as a processing flow that is automatically executed in an electronic device, so that the processing flow can be directly called when the correlation coefficient needs to be calculated, and the corresponding relationship is determined based on each first parameter within the candidate first parameter range of each dimension, and then the first target parameters of each dimension whose corresponding relationship meets the target screening condition are taken as parameters in the first target parameter group. In one embodiment, when the corresponding relationship is represented by the correlation coefficient of the real-time processing parameter and the target processing parameter, the target screening condition can be a condition that represents the maximum correlation coefficient. When determining each first parameter, the first parameter can be randomly determined according to its corresponding first parameter range to calculate the correlation coefficient and screen it. It can also be that the first parameter is traversed according to the priority of the dimension.

[0110] Correspondingly, in one implementation of the embodiments of the present application, the corresponding relationships among the first parameters within the candidate first parameter range of each dimension are determined, including: based on the priority of each dimension, traversing each first parameter within the candidate first parameter range of each dimension in turn to obtain the corresponding relationships among the first parameters under each dimension.

[0111] Among them, the priority of the dimension of the first parameter group can be determined based on the calculation priority corresponding to the controller. Different types of controllers have different dimensions of the first parameter group, and the priority of the corresponding dimensions may also be different. For example, if the controller is a PID controller, the priority order of proportion (P), integration (I), and differentiation (D) is determined according to the importance of the parameters in the PID algorithm, and then the first parameters in the candidate first parameter range are traversed in turn according to the priority to determine the corresponding correspondence. Specifically, the candidate first parameter range can be determined based on the size of the current real-time processing parameter and the target processing parameter, that is, the first parameter range can include a parameter range that controls the increase of the second parameter, or a parameter range that controls the decrease of the second parameter.

[0112] In one embodiment, based on the priority of each dimension, each first parameter within the candidate first parameter range of each dimension is traversed in turn to obtain corresponding relationships corresponding to the first parameters in each dimension; based on the target screening conditions corresponding to the several corresponding relationships corresponding to each first parameter, the first target parameter of each dimension is determined from the candidate first parameter range in each dimension, including: determining the target dimension currently traversed, keeping the first parameters of other dimensions unchanged, wherein the first parameters of the traversed dimensions in other dimensions are configured as the first target parameters determined under the dimension; traversing each first parameter under the target dimension so that the controller generates multiple corresponding second parameters in turn; based on the real-time processing parameters and target processing parameters of the target task corresponding to each second parameter, the corresponding relationships corresponding to each first parameter under the target dimension are obtained; based on the target screening conditions corresponding to the several target relationships corresponding to each first parameter under the target dimension, the first target parameter of the target dimension is determined; traversing other dimensions in turn according to the above steps until the first target parameters under all dimensions are determined.

[0113] The target screening condition is to screen out the conditions that are most relevant to the real-time processing parameters and the target processing parameters that characterize the corresponding relationship. For example, the corresponding relationship is determined by the deviation values ​​of the real-time processing parameters and the target processing parameters corresponding to each first parameter. The target screening condition can be the condition that characterizes the smallest deviation value. Taking the corresponding relationship as a correlation coefficient as an example, the target screening condition can be the condition with the largest correlation coefficient.

[0114] Depending on the type of controller, the parameter dimensions corresponding to the controller are different, and further, the priorities of each dimension may also be different. Taking the PID controller, the dimensions are proportional (P), integral (I), and differential (D) as an example. According to the PID algorithm, the priority order of the corresponding dimensions is: proportional (P), integral (I), and differential (D). For another example, the controller is a thread controller, and the dimensions may include the thread quantity control dimension and the thread processing capacity dimension. The corresponding dimension priority may be that the thread processing capacity dimension has a higher priority than the thread quantity control dimension, that is, when adjusting the processing thread, first adjust the relevant parameters of the thread processing capacity, and then adjust the number of available threads, etc.

[0115] In a specific embodiment, taking the controller as a PID controller as an example, the proportional parameters corresponding to the P dimension can be traversed first, and the I1 and D1 parameters in the I and D dimensions can be used as fixed parameters. Specifically, any parameter in the I and D dimensions can be randomly selected as a fixed parameter. In order to further improve the processing efficiency, the parameters of the I and D dimensions determined in the previous monitoring cycle for the target task can also be used as the fixed parameters of the current traversal, or the parameters commonly used in the historical processing of the I and D dimensions can be used as the current fixed parameters. Then, according to the P1 parameter in the current range of the P dimension, the first parameter group (P1, I1, D1) at this time is determined, and the parameter group is used as the parameter of the PID controller to determine the second parameter U1. After adjusting the scheduling resources of the target task based on the second parameter U1, the corresponding real-time processing parameter X1 and the target processing parameter Y1 are obtained, and then the corresponding correlation coefficient C11 is calculated.

[0116] Then, according to the current P2 parameter of the P dimension, a parameter group (P2, I1, D1) is determined, and the parameter group is used as the parameter of the PID controller to determine the second parameter U2. After adjusting the scheduling resources of the target task based on the second parameter U2, the corresponding real-time processing parameter X2 and the target processing parameter Y2 are obtained within the sampling period, and then the corresponding correlation coefficient C12 is calculated. It should be noted that the target processing parameters Y1 and Y2 can be the same or different. For example, if the load or data transmission conditions or operating state of the electronic device changes, the target processing parameter can also change accordingly to ensure the performance of the current electronic device. Until all the current candidate parameters of the P dimension are traversed, several correlation coefficients are obtained. For example, there are a total of m candidate parameters in the P dimension, and a total of m traversals are performed, then m correlation coefficients will be obtained, that is, correlation coefficients C11, C12, ... C1m can be obtained. The parameters corresponding to the maximum correlation coefficient in the current dimension can be determined based on these correlation coefficients as the default parameters of the dimension. For example, the default parameter of the P dimension can be P3. Then traverse the I dimension again. The traversal method here is the same as the traversal method of the P dimension. At this time, the parameters of the P dimension are fixed to P3, and the D dimension is fixed to D1. If the parameters are traversed n times in the I dimension, n correlation coefficients will be obtained, that is, the correlation coefficients C21, C22, ... C2n will be obtained. Then, based on these correlation coefficients, the parameters of the I dimension corresponding to the maximum correlation coefficient are determined as the default parameters of the I dimension, such as I4. Finally, traverse the D dimension again. The traversal method is the same as above. At this time, the parameters of the P dimension are fixed to P3, and the I dimension is fixed to I4. After w traversals in the D dimension, w correlation coefficients will be obtained, that is, w correlation coefficients will be obtained, that is, correlation coefficients C31, C32, ..., C3w will be obtained. Then, based on these correlation coefficients, the parameters of the D dimension corresponding to the maximum correlation coefficient are determined as the default parameters of the D dimension, such as the obtained parameters of the D dimension are D5. Based on the above determination of the parameter group (P3, I4, D5), the parameter group (P3, I4, D5) is determined as the first target parameter group.

[0117] After obtaining the first target parameter group, the PID controller will calculate the current second target parameter based on the (P3, I4, D5) parameters, and control the scheduling of resources for the execution of the target task according to the second target parameter. For example, the processing thread of the target task can be controlled based on the second target parameter to obtain more scheduling resources to improve its processing efficiency, so that the real-time processing parameters of the target task are within the matching range of the target processing parameters, thereby ensuring that the electronic device is within a better energy consumption ratio range when executing the target task, meeting the actual application scenario requirements.

[0118] In the above embodiment, the process of determining the first target parameter group is described by taking the corresponding relationship through the correlation coefficient as an example. When the corresponding relationship is represented by other coefficients, the traversal process of determining the first target parameter group is similar to the above, and will not be described in detail here.

[0119] Correspondingly, a performance adjustment method is also provided in the embodiment of the present application, see Figure 4 , the method may include the following steps:

[0120] S401: Obtain current real-time processing parameters of a target task executed by an electronic device.

[0121] S402: In response to the real-time processing parameter satisfying the update condition, update the second target parameter based on the first target parameter group.

[0122] S403: Control the scheduling of execution resources of the target task according to the updated second target parameter.

[0123] The first target parameter group is determined based on the correspondence between the real-time processing parameters and the target processing parameters. Specifically, the determination process of the first target parameter group can refer to the description of the control parameter determination method in the above embodiment, which will not be described in detail here.

[0124] In the case where the real-time processing parameter meets the update condition, the second target parameter can be updated based on the first target parameter group. The update condition can be determined based on the current application scenario. For example, in an application scenario with high processing performance requirements, the update condition can be determined based on the monitoring cycle, that is, if the processing performance of the target task corresponding to the real-time processing parameter is monitored to be lower than the performance condition in each monitoring cycle, it is determined that the real-time processing parameter meets the update condition. The update condition can also be determined based on the operating state of the electronic device. For example, the update condition represents the condition that the power consumption of the electronic device is lower than the power consumption threshold, that is, when it is detected that the electronic device runs the target task based on the real-time processing parameter, the power consumption of the electronic device is lower than the power consumption threshold, which means that the real-time processing parameter meets the update condition. For another example, the update condition is related to the pre-set target processing parameter, and the real-time processing parameter meeting the update condition can represent that there is a certain deviation between the real-time processing parameter and the target processing parameter.

[0125] Specifically, in the process of updating the second target parameter based on the first target parameter group, the processing thread corresponding to the target task can be obtained. The processing thread can be a thread that performs data processing when the target task is running, or a thread that satisfies resource scheduling when the target task is running, that is, the processing thread is a thread that affects the generation of real-time processing parameters corresponding to the target task. The updated second target parameter is used to control and adjust the working parameters of the processing thread so that the working parameters are adjusted to the corresponding parameter values. Furthermore, the minimum value of the working parameter of the processing thread can also be adjusted to the parameter value of the working parameter corresponding to the above-mentioned second target parameter. Thereby, the energy consumption ratio condition when the target task is running is met, and the effect of the target task processing is guaranteed.

[0126] The following takes the application scenario of applying the above-mentioned control parameter determination method to the dynamic performance framework ADPF of the Android operating system as an example to illustrate the performance adjustment method in the embodiment of the present application. Android DynamicPerformance Framework (ADPF) is an Android dynamic performance framework that provides a series of interfaces for games and performance-sensitive applications to call. Using these interfaces, you can dynamically monitor thermal and dynamically adjust the resources you need according to thermal and your own needs. Among them, thermal is an important component of ADPF, which represents the monitoring of the thermal state of the application, which allows the thermal state of the application to dynamically adjust the performance to avoid overheating and performance throttling.

[0127] See also Figure 5 In the rendering and display processing of the Android system, there will be Vsync (Vertical Synchronization, used to synchronize the GPU frame rendering with the refresh rate of the display to avoid screen tearing and freezing) related information in the rendering component (such as the Surface Flinger component in the Android system) and the display component (such as the HWUI component in the Android system). The rendering component or the display component will continuously upload the actual rendering time and expected rendering time of a frame of image of Vsync, and pass this information to the performance optimization component of the electronic device, such as the performance optimization component can be PowerHal. PowerHal is a key component for power management and performance optimization in the Android system. It enables the system to dynamically adjust the behavior of the hardware according to the current workload and power requirements to provide the best performance and power efficiency.

[0128] The performance optimization component will calculate and output the second target parameter using the PID parameter in the first target parameter group through the PID algorithm. In this application scenario, the second target parameter can be a performance adjustment value, such as expressed by the Uclamp value. The main rendering thread is set to this Uclamp value to dynamically adjust the resource occupancy of SurfaceFlinger / HWUI (rendering and display). Among them, Uclamp (Utilization Clamping) is a scheduler feature in the Linux kernel, which allows user space to help manage the performance requirements of tasks. Uclamp sets the performance range of the task by providing two adjustable parameters: Uclamp_min and Uclamp_max, thereby ensuring that the task runs within this range. Specifically, the Uclamp_min of the main thread can be set to the adjustment value calculated by the PID algorithm. Thereby ensuring the energy consumption ratio and display effect of rendering and display.

[0129] Taking the application scenario of data cache as an example, the target task is the data cache task, the processing parameter represents the data cache amount per unit time, the real-time processing parameter can represent the data cache amount of the current data cache task per unit time (such as referred to as the first cache parameter), and the target processing parameter can represent the data cache amount of the expected data cache task per unit time (such as referred to as the second cache parameter).

[0130] In order to ensure the data caching effect and the power consumption of the electronic device when performing the data caching task, that is, to maintain the electronic device in a state of better energy consumption ratio, the corresponding second cache parameter can be set according to the operating state of the electronic device. That is, during the operation of the data caching task, the second cache parameter can be different in different operating stages. For example, the second cache parameter can be determined according to the operating load of the electronic device. If the current electronic device has other processing tasks in addition to performing data storage tasks, it means that the operating load of the electronic device is relatively high. Correspondingly, in order to ensure the normal energy consumption of the electronic device, the second cache parameter set at this time can be set to the first parameter; if the current operating load of the electronic device is relatively low, the second cache parameter can be set to the second parameter, wherein the parameter value of the first parameter is less than the parameter value of the second parameter.

[0131] In this scenario, the corresponding relationship between the first cache parameter and the second cache parameter is characterized by the correlation coefficient between the two. Correspondingly, the calculation process of the correlation coefficient can refer to the calculation process of formula (4), that is, the correlation coefficient is a coefficient after the initial correlation coefficient is corrected based on the correction coefficient. At different stages of executing the data cache task, the selected correction coefficient may be different. For example, in the early stage of executing the data cache task, it is necessary to ensure the performance of the data cache. At this time, performance is prioritized, and the correction coefficient can select a relatively small value; correspondingly, in the later stage of executing the data cache task, in order to ensure the normal operation of the data cache task and the power consumption of the electronic device, the correction coefficient can select a relatively large value. This can meet the task requirements of different operating stages and improve the processing effect of the data cache task.

[0132] Furthermore, the target processing parameter (i.e., the second cache parameter) and the correction coefficient for calculating the correlation coefficient can also be updated simultaneously in different execution stages of the data cache task or in accordance with the task requirements for real-time adjustment. For example, when the electronic device is in a state of high operating power consumption, in order to ensure the smooth caching of data, the second cache parameter can be set to a relatively small value, and the correction parameter can be adjusted synchronously to ensure the efficiency of data caching. After the correlation coefficient is determined based on the current application scenario characteristics and requirements, the control parameter for updating the real-time processing parameter (i.e., the first cache parameter) can be determined based on the correlation coefficient with reference to the aforementioned embodiment, and then the processing thread or processing resource affecting the real-time processing parameter can be adjusted based on the control parameter to ensure the data caching effect and the operating status of the electronic device. The specific parameter update process can be referred to the aforementioned embodiment, which will not be repeated here.

[0133] See also Figure 6 , is a flow chart of a method for determining PID parameters provided in an embodiment of the present application. The method may include the following steps:

[0134] S601. Obtain current PID parameters.

[0135] S602: Determine whether the current PID parameters meet the conditions for ending the traversal, if yes, end the traversal. If no, execute step S603.

[0136] S603, executing the automated traversal script to calculate and obtain the correlation coefficient.

[0137] The automated traversal script refers to an operation script that obtains the current real-time processing parameters and target processing parameters, and calculates the correlation coefficient between the two based on the target traversal method. Figure 5In the application scenario shown, the real-time processing parameter represents the actual time to render a frame of image, and the target processing parameter represents the expected time to render a frame of image. The specific traversal method is based on the importance of the PID algorithm, that is, the order of proportion (P), integration (I), and differentiation (D). Each parameter is traversed in turn, and the correlation coefficient Corr corresponding to each group of parameters is calculated. The specific processing process can be found in the description of the aforementioned embodiment, which will not be described in detail here.

[0138] Corresponding to the processing process of the PID controller generating the second parameter, the dimension of the first parameter group can also be refined into the proportional, integral, and differential dimensions that can control the increase of the second parameter (for example, control the second parameter to become larger), and the proportional and differential dimensions that can control the decrease of the second parameter. It can also include parameter dimensions for determining how much data to take when calculating the second parameter value (the parameter dimension can also include corresponding proportional, integral, and differential dimensions). It can be seen that in the dimensional mode, different parameters correspond to parameters in multiple dimensions. For example, the proportional parameter corresponds to proportional parameter values ​​in multiple dimensions, specifically including the proportional parameter value that controls the increase of the second parameter, the proportional parameter value that controls the decrease of the second parameter, and the proportional parameter value of the parameter dimension used to determine how much data to take when calculating the second parameter value. In each proportional parameter, it is also traversed according to the corresponding priority, so that the corresponding parameter can be obtained quickly and accurately.

[0139] For example, in the PID controller, the parameter dimensions that can control the second parameter to rise are represented by Po, I and Do, that is, the process of controlling the second parameter to rise can be achieved by adjusting the relevant parameters of the proportional dimension (P), integral dimension (I) and differential dimension (D). Correspondingly, the parameters that control the second parameter to fall are represented by Pu and Du, that is, the process of controlling the second parameter to fall can be achieved by adjusting the parameters of the proportional dimension (P) and differential dimension (D). The process of controlling the output of the PID controller can also include parameter dimensions for determining how much data to take when calculating the value of the second parameter, such as the data step sub-dimension and the data sampling sub-dimension. When controlling this dimension, it can also be achieved based on adjusting the proportional dimension (P), integral dimension (I) and differential dimension (D).

[0140] In the above application process, the priority of the purpose dimension corresponding to each control purpose can be determined first, and then traversal can be performed according to each parameter dimension under each purpose dimension. For example, in the current application scenario, it is necessary to first increase the second parameter. At this time, the priority of the purpose dimension for controlling the increase of the second parameter is higher than the priority of the purpose dimension for controlling the decrease of the second parameter. Then, under the purpose dimension for controlling the increase of the second parameter, traversal can be performed according to the priority dimensions of P, I, and D. After obtaining the PID parameter under the purpose dimension for controlling the second parameter, the specific traversal process can be referred to the description of the above embodiment, which will not be described in detail here. Then traverse the purpose dimension for controlling the decrease of the second parameter, and then traverse according to the priority dimensions of P, I, and D under the purpose dimension to obtain the corresponding parameter. Correspondingly, it is also possible to first use the order of priority of P, I, and D as the first-level traversal order, and then use the purpose dimension under the parameter of the corresponding dimension as the second-level traversal order for traversal. Specifically, if the data is traversed with the priority of P, I, and D, the data under the P dimension is traversed first. When traversing the data of the P dimension, it is necessary to traverse according to the priority of the target dimension corresponding to the control purpose. For example, when traversing the data of the P dimension, the data is first traversed with the purpose of controlling the second parameter to rise. After the traversal of the target dimension is completed, the P dimension is obtained. k , the P k It is characterized that when the traversal is performed with the purpose of controlling the second parameter to rise, the correlation coefficient between the real-time processing parameter corresponding to the second parameter obtained and the target processing parameter is the largest. Then the traversal is performed with the purpose of controlling the second parameter to fall, and then according to the screening condition with the largest relative relationship, P is obtained at this time. w , P w Characterize that when traversing for the purpose of controlling the decrease of the second parameter, according to P w The correlation coefficient between the real-time processing parameter and the target processing parameter corresponding to the second parameter is the largest. k and P w The corresponding correlation coefficients are: k The corresponding correlation coefficient is the largest, so the parameter taken in the P dimension is P k Then, when traversing the data of dimension I, the data is traversed according to the priority of the target dimension corresponding to the control purpose. For example, the data is traversed with the purpose of controlling the second parameter to rise first. After the traversal of the target dimension is completed, the data is traversed with the purpose of controlling the second parameter to fall. Then, the target parameter of dimension I is determined according to the screening conditions of the parameters. Similarly, the target parameter of dimension D is obtained. In this way, the traversal process can be combined with the control purpose, making the traversal process more efficient and accurate.

[0141] S604: The parameter with the largest correlation coefficient in each dimension is used as the default parameter of the dimension, and the next traversal is performed. The above steps S601 to S604 are repeatedly performed until the final PID parameter is obtained, and the traversal is terminated.

[0142] In the embodiment of the present application, the relationship between the real-time processing parameters and the target processing parameters is analyzed, and the optimal PID parameters are determined by means of an automated script. There is no need to manually debug and determine the PID parameters, which can save time in determining the PID parameters. The correlation coefficient can also be used as a criterion to determine the quality of the PID parameters, making the parameter determination process more objective and accurate, and ensuring the performance of the electronic device in performing the target task. In the process of calculating the correlation coefficient in the embodiment of the present application, a correction coefficient is introduced when the real-time processing parameter is greater than the target processing parameter, achieving a balance between performance and power consumption, and meeting the needs of different application scenarios by adjusting the correction coefficient.

[0143] In another embodiment of the present application, a device for determining control parameters of an electronic device is also provided. Figure 7 , the device comprises:

[0144] A first acquisition unit 701 is used to obtain real-time processing parameters and target processing parameters of a target task executed by the electronic device;

[0145] A second acquisition unit 702 is used to acquire a corresponding relationship between the real-time processing parameter and the target processing parameter;

[0146] A first determining unit 703 is configured to determine at least one first target parameter group based on the corresponding relationship, where the first target parameter group includes at least one first target parameter;

[0147] The second determination unit 704 is configured to determine a second target parameter based on the first target parameter group, wherein the second target parameter is related to the execution resource scheduling of the target task.

[0148] Obtaining real-time processing parameters and target processing parameters of a target task performed by the electronic device;

[0149] Acquiring a corresponding relationship between the real-time processing parameter and the target processing parameter;

[0150] Based on the corresponding relationship, determining at least one first target parameter group, wherein the first target parameter group includes at least one first target parameter;

[0151] Based on the first target parameter group, a second target parameter is determined, wherein the second target parameter is related to the execution resource scheduling of the target task.

[0152] Optionally, the corresponding relationship between the real-time processing parameter and the target processing parameter includes: a correlation coefficient between the real-time processing parameter and the target processing parameter;

[0153] The correlation coefficient is calculated based on the real-time processing parameter obtained by sampling at a target sampling number and the target processing parameter.

[0154] Optionally, the corresponding relationship between the real-time processing parameter and the target processing parameter includes: a correlation coefficient between the real-time processing parameter and the target processing parameter;

[0155] The method for forming the correlation coefficient includes: obtaining an initial correlation coefficient by calculating the real-time processing parameter obtained by sampling at a target sampling number and the target processing parameter; and obtaining the correlation coefficient by correcting the initial correlation coefficient based on a correction coefficient.

[0156] Optionally, when the real-time processing parameter is greater than the target processing parameter, the correction coefficient is obtained by at least calculating based on a deviation value between the real-time processing parameter and the target processing parameter.

[0157] Optionally, the first determining unit includes:

[0158] A first acquisition subunit, configured to obtain a dimension of a first parameter group of a controller, wherein the controller is configured to generate a second parameter based on the first parameter group; the first parameter group includes at least one first parameter;

[0159] A second determining subunit is used to determine the corresponding relationships between the first parameters in the candidate first parameter range of each dimension;

[0160] A second determination subunit is used to determine the first target parameter of each dimension from the candidate first parameter range in each dimension based on the target screening conditions corresponding to the plurality of corresponding relationships corresponding to the first parameters;

[0161] The combining subunit is used to take the combination of the first target parameters of each dimension as the first target parameter group.

[0162] Optionally, the second determining subunit is configured to:

[0163] Based on the priority of each dimension, each first parameter within the candidate first parameter range of each dimension is traversed in turn to obtain the corresponding relationships between the first parameters in each dimension.

[0164] Optionally, based on the priority of each dimension, each first parameter within the candidate first parameter range of each dimension is traversed in turn to obtain corresponding relationships corresponding to the first parameters in each dimension; based on the target screening conditions corresponding to the corresponding relationships corresponding to the first parameters, the first target parameter of each dimension is determined from the candidate first parameter range in each dimension, including:

[0165] Determine the target dimension of the current traversal, and keep the first parameters of other dimensions unchanged, wherein the first parameters of the traversed dimensions among the other dimensions are configured as the first target parameters determined under the dimension;

[0166] Traversing each first parameter under the target dimension, so that the controller generates a corresponding plurality of second parameters in sequence;

[0167] Based on the real-time processing parameters and target processing parameters of the target task corresponding to each of the second parameters, the corresponding relationships corresponding to each of the first parameters under the target dimension are obtained;

[0168] Determine the first target parameter of the target dimension based on the target screening conditions corresponding to the plurality of corresponding relationships corresponding to the first parameters under the target dimension;

[0169] According to the above steps, other dimensions are traversed in sequence until the first target parameters in all dimensions are determined.

[0170] Optionally, the controller comprises a PID controller.

[0171] Correspondingly, a performance adjustment device is also provided in the embodiment of the present application, see Figure 8 , the device may include:

[0172] The third acquisition unit 801 is used to obtain the current real-time processing parameters of the target task executed by the electronic device;

[0173] An updating unit 802, configured to update a second target parameter based on a first target parameter group in response to the real-time processing parameter satisfying an update condition; wherein the first target parameter group is determined based on a correspondence between the real-time processing parameter and the target processing parameter;

[0174] The control unit 803 is used to control the scheduling of the execution resources of the target task according to the updated second target parameter.

[0175] It should be noted that the specific implementation of each unit and sub-unit in this embodiment can refer to the corresponding content in the previous text and will not be described in detail here.

[0176] In another embodiment of the present application, a readable storage medium is also provided, on which a computer program is stored. When the computer program is executed by a processor, it implements the control parameter determination method of the electronic device as described above; when the computer program is executed by the processor, it can also implement the performance adjustment method as described above.

[0177] In another embodiment of the present application, an electronic device is further provided, and the electronic device may include:

[0178] A memory, used to store applications and data generated by the operation of the applications;

[0179] The processor is used to execute the application program to implement the control parameter determination method of the electronic device and / or implement the above-mentioned performance adjustment method.

[0180] It should be noted that the specific implementation of the processor in this embodiment can refer to the corresponding content in the previous text and will not be described in detail here.

[0181] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0182] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in the above description according to function. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0183] The steps of the method or algorithm described in conjunction with the embodiments disclosed herein may be implemented directly using hardware, a software module executed by a processor, or a combination of the two. The software module may be placed in a random access memory (RAM), a memory, a read-only memory (ROM), an electrically programmable ROM, an electrically erasable programmable ROM, a register, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.

[0184] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for determining a control parameter of an electronic device, comprising: Obtaining real-time processing parameters and target processing parameters of a target task performed by the electronic device; Acquiring a corresponding relationship between the real-time processing parameter and the target processing parameter; Based on the corresponding relationship, determining at least one first target parameter group, wherein the first target parameter group includes at least one first target parameter; Based on the first target parameter group, a second target parameter is determined, wherein the second target parameter is related to the execution resource scheduling of the target task.

2. According to the method of claim 1, the correspondence between the real-time processing parameter and the target processing parameter comprises: a correlation coefficient between the real-time processing parameter and the target processing parameter; The correlation coefficient is calculated based on the real-time processing parameter obtained by sampling at a target sampling number and the target processing parameter.

3. According to the method of claim 2, the corresponding relationship between the real-time processing parameter and the target processing parameter comprises: a correlation coefficient between the real-time processing parameter and the target processing parameter; The method for forming the correlation coefficient includes: obtaining an initial correlation coefficient by calculating based on the real-time processing parameter obtained by sampling at a target sampling number and the target processing parameter; The correlation coefficient is obtained by correcting the initial correlation coefficient based on the correction coefficient.

4. The method according to claim 3, wherein when the real-time processing parameter is greater than the target processing parameter, the correction coefficient is obtained by at least calculating based on a deviation value between the real-time processing parameter and the target processing parameter.

5. The method according to claim 1, 2, 3 or 4, wherein determining at least one first target parameter group based on the corresponding relationship comprises: obtaining a dimension of a first parameter set of a controller, the controller being configured to generate a second parameter based on the first parameter set; The first parameter group includes at least one first parameter; Determine the corresponding relationships between the first parameters within the candidate first parameter range of each dimension; Based on the target screening conditions corresponding to the plurality of corresponding relationships corresponding to the first parameters, determine the first target parameter of each dimension from the candidate first parameter range in each dimension; A combination of the first target parameters in various dimensions is taken as the first target parameter group.

6. The method according to claim 5, wherein determining the corresponding relationship between each first parameter in the candidate first parameter range of each dimension comprises: Based on the priority of each dimension, each first parameter within the candidate first parameter range of each dimension is traversed in turn to obtain the corresponding relationships between the first parameters in each dimension.

7. The method according to claim 6, based on the priority of each dimension, traversing each first parameter in the candidate first parameter range of each dimension in turn to obtain the corresponding relationships corresponding to the first parameters in each dimension; based on the target screening conditions corresponding to the corresponding relationships corresponding to the first parameters, determining the first target parameters of each dimension from the candidate first parameter range in each dimension, comprising: Determine the target dimension of the current traversal, and keep the first parameters of other dimensions unchanged, wherein the first parameters of the traversed dimensions among the other dimensions are configured as the first target parameters determined under the dimension; Traversing each first parameter under the target dimension, so that the controller generates a corresponding plurality of second parameters in sequence; Based on the real-time processing parameters and target processing parameters of the target task corresponding to each of the second parameters, the corresponding relationships corresponding to each of the first parameters under the target dimension are obtained; Determine the first target parameter of the target dimension based on the target screening conditions corresponding to the plurality of corresponding relationships corresponding to the first parameters under the target dimension; According to the above steps, other dimensions are traversed in sequence until the first target parameters in all dimensions are determined.

8. The method of claim 5, wherein the controller comprises a PID controller.

9. A performance adjustment method, comprising: Obtaining current real-time processing parameters of a target task executed by an electronic device; In response to the real-time processing parameter satisfying an update condition, updating the second target parameter based on the first target parameter group; wherein the first target parameter group is determined based on a correspondence between the real-time processing parameter and the target processing parameter; The scheduling of the execution resources of the target task is controlled according to the updated second target parameter.

10. A device for determining a control parameter of an electronic device, comprising: A first acquisition unit, used to obtain real-time processing parameters and target processing parameters of a target task executed by the electronic device; A second acquisition unit, used for acquiring a correspondence between the real-time processing parameter and the target processing parameter; A first determining unit, configured to determine at least one first target parameter group based on the corresponding relationship, wherein the first target parameter group includes at least one first target parameter; The second determination unit is used to determine a second target parameter based on the first target parameter group, wherein the second target parameter is related to the execution resource scheduling of the target task.

11. A performance adjustment device, comprising: A third acquisition unit, used to obtain the current real-time processing parameters of the target task executed by the electronic device; An updating unit, configured to update a second target parameter based on a first target parameter group in response to the real-time processing parameter satisfying an update condition; wherein the first target parameter group is determined based on a correspondence between the real-time processing parameter and the target processing parameter; A control unit is used to control the scheduling of execution resources of the target task according to the updated second target parameter.

Images