Back to results
US 9,791,911 B2Granted
Determining whether a change in power usage is abnormal when power usage exceeds a threshold based on additional metrics of components in an electronic device
AssigneeChiun Mai Communication Systems, Inc.
InventorsChia-Jung Hu, Kai-Hsi Yang, Tai-Chuan Chen, Li-Bin Shen
Life of this patent
Estimating the term…
Abstract
In a method for detecting power usages of an electronic device, parameters of components of the electronic device and a power usage of the electronic device are obtained within a predetermined Nth period. The electronic device determines if the power usage of the electronic device is normal according to the parameters of components of the electronic device within the Nth period. The electronic device records information of applications which run in the electronic device within the Nth period when the power usage of the electronic device is abnormal.
Claims14
An electronic device comprising:
- at least one processor;
- and
- a storage device that stores one or more programs, which when executed by the at least one processor, cause the at least one processor to:
- obtain parameters of components of the electronic device and a power usage of the electronic device within a predetermined Nth period, wherein N is a positive integer;
- determine if the power usage of the electronic device is normal according to the parameters of components of the electronic device within the Nth period when a difference between the power usage of the electronic device within the Nth period and a power usage of the electronic device within a (N−1)th period exceeds a threshold value;
- and
- record, when the power usage of the electronic device is abnormal, information of applications running on the electronic device within the Nth period;
- wherein each of the components of the electronic device is assigned with a weight factor for determining whether the power usage of the electronic device is normal or abnormal;
- wherein the weight factor of each of the components is calculated by performing operations comprising:
- saving the power usage of the electronic device and the parameters of the components into a list at each of a plurality of moments;
- and
- when a difference between the power usage of a first one of the moments and the power usage of a second one of the moments is greater than a predetermined value, marking the power usage and the parameters of the second moment as error values and removing the error values from the list, wherein the second moment immediately follows the first moment;
- and
- calculating the weight factor of each of the components according to the list using a multivariate nonlinear regression algorithm and a recursive algorithm.
11 more elements
A computer-based method for detecting power usages of an electronic device, the method comprising:
- obtaining parameters of components of the electronic device and a power usage of the electronic device within a predetermined Nth period, wherein N is a positive integer;
- determining if the power usage of the electronic device is normal according to the parameters of components of the electronic device within the Nth period when a difference between the power usage of the electronic device within the Nth period and a power usage of the electronic device within a (N−1)th period exceeds a threshold value;
- and
- recording, when the power usage of the electronic device is abnormal, information of applications running on the electronic device within the Nth period;
- wherein each of the components of the electronic device is assigned with a weight factor for determining whether the power usage of the electronic device is normal or abnormal;
- wherein the weight factor of each of the components is calculated by performing operations comprising:
- saving the power usage of the electronic device and the parameters of the components into a list at each of a plurality of moments;
- when a difference between the power usage of a first one of the moments and the power usage of a second one of the moments is greater than a predetermined value, marking the power usage and the parameters of the second moment as error values and removing the error values from the list, wherein the second moment immediately follows the first moment;
- and
- calculating the weight factor of each of the components according to the list using a multivariate nonlinear regression algorithm and a recursive algorithm.
7 more elements
A non-transitory computer-readable medium having stored thereon instructions that, when executed by at least one processor of an electronic device, causing the processor to perform a method for detecting power usages of the electronic device, the method comprising:
- obtaining parameters of components of the electronic device and a power usage of the electronic device within a predetermined Nth period, wherein N is a positive integer;
- determining if the power usage of the electronic device is normal according to the parameters of components of the electronic device within the Nth period when a difference between the power usage of the electronic device within the Nth period and a power usage of the electronic device within a (N−1)th period exceeds a threshold value;
- and
- recording, when the power usage of the electronic device is abnormal, information of applications running on the electronic device within the Nth period;
- wherein each of the components of the electronic device is assigned with a weight factor for determining whether the power usage of the electronic device is normal or abnormal;
- wherein the weight factor of each of the components is calculated by performing operations comprising:
- saving the power usage of the electronic device and the parameters of the components into a list at each of a plurality of moments;
- when a difference between the power usage of a first one of the moments and the power usage of a second one of the moments is greater than a predetermined value, marking the power usage and the parameters of the second moment as error values and removing the error values from the list, wherein the second moment immediately follows the first moment;
- and
- calculating the weight factor of each of the components according to the list using a multivariate nonlinear regression algorithm and a recursive algorithm.
7 more elements
Dependencies are inferred from the claim text.