TWI717878B - Battery power prediction method and computer program product thereof - Google Patents

Abstract

A method for predicting battery power, suitable for a carriable device, comprising: detecting one or more applications currently executing on the mobile device and obtaining names of the one or more applications; inputting the names of the one or more applications to a battery power prediction model to calculate a battery power drop value per unit time, wherein the battery power prediction model is trained to store an equivalent power consumption weight corresponding to the one or more applications; the battery power prediction model correspondingly finds the equivalent power consumption weight of the one or more applications according to different names of the one or more applications that are input; and determining whether to issue a battery warning according to the battery power drop value per unit time.

Description

Method for predicting battery power and its computer program product

The present invention relates to a method for predicting battery power, and more particularly, to a method for predicting battery power for portable devices and computer program products.

Current electronic devices only display the current percentage of battery power, but they do not tell the user the impact of the next program on the remaining battery power, or whether it is necessary to connect a power cord or carry a power cord for the next trip.

The problem with the current display of the remaining battery power is that it does not refer to the user's usage habits at different times. More specifically, for example, the battery level is displayed for 5 hours at 3 in the afternoon, but when the user is playing a game at 5:30 in the afternoon, some programs in the generated process tree will enable graphics processing The GPU (Graphics Processing Unit) causes higher power consumption. Perhaps the device shuts down before 6 o'clock in the afternoon because the battery runs out.

The method for predicting battery power according to an embodiment of the present invention is applicable to a portable device that uses a battery power prediction model carried in it to determine whether to issue a power warning. The prediction method Including: detecting one or more applications currently running on the portable device, and obtaining the names of the one or more applications; inputting the names of the one or more applications into the battery power prediction model for Calculate a value of the decrease in battery power per unit time; wherein the battery power prediction model is trained to store equivalent power consumption weights corresponding to the one or more applications; the battery power prediction model is based on the input of the one or more applications According to the difference of the name, find out the equivalent power consumption weight of the one or more applications accordingly; determine whether to issue the power warning according to the value of the battery power decrease per unit time.

The method for predicting battery power as described above further includes: detecting the current time point to determine whether the current time point falls within an early warning period before a scheduled event occurs; wherein, the current time when the scheduled event occurs , The portable device will execute the one or more applications; when the current time falls within the warning period before the scheduled event, enter the name of the one or more applications into the battery level The prediction model predicts the battery power decrease value per unit time when the scheduled event occurs; according to the battery power decrease value per unit time, it is judged whether to issue another power warning.

The battery power prediction method described above, wherein the training method of the battery power prediction model includes: detecting a total power consumption value per unit time when the one or more applications are running on the portable device; adjusting the The equivalent power consumption weights of one or more applications, so that the battery power prediction model calculated based on the equivalent power consumption weights of the one or more applications, the battery power decline value per unit time can be close to the total unit time Power consumption value; repeatedly adjust the equivalent power consumption weights of one or more applications until the difference between the battery power decrease value per unit time and the total power consumption value per unit time is less than or equal to a threshold; when the battery power per unit time If the difference between the decrease value and the total power consumption value per unit time is less than or equal to the threshold, the current equivalent power consumption weight is set and stored as the equivalent power consumption weight corresponding to the one or more applications.

Such as the method for predicting battery power, wherein the power warning is a display screen of the portable device to display a message that a power cord needs to be connected.

Such as the method for predicting battery power, wherein the other power warning is to display a message that a power cord needs to be carried through a display screen of the portable device.

The method for predicting battery power as described above, wherein the decrease in battery power per unit time is the power consumption of the battery per hour, or the percentage of power that the battery decreases per hour.

The computer program product according to an embodiment of the present invention is suitable for a portable device to determine whether to issue a power warning. The computer program product is loaded into the program by the computer to execute: a first detection command to make the A processor of the computer detects one or more application programs currently running on the portable device, and obtains the name of the one or more application programs; a call instruction causes the processor to obtain from the computer's memory A battery power prediction model; wherein the battery power prediction model is trained to store equivalent power consumption weights corresponding to the one or more application programs; the battery power prediction model is based on the input name of the one or more application programs , And correspondingly find out the equivalent power consumption weights of the one or more applications; a first input command causes the processor to input the names of the one or more applications into the battery power prediction model to calculate a A value of the decrease in battery power per unit time; a first judgment command enables the processor to determine whether to issue the power warning according to the value of the decrease in battery power per unit time.

For example, the above-mentioned computer program product further includes: a second detection command to enable the processor to detect the current time point to determine whether the current time point falls within an early warning period before a scheduled event occurs; Wherein, at the moment when the scheduled event occurs, the portable device will execute the one or more applications; a second input command, when the current point in time falls within the warning period before the scheduled event occurs , Enabling the processor to input the names of the one or more application programs into the battery power prediction model to predict the decrease in battery power per unit time when the scheduled event occurs; a second judgment instruction enables the processor to follow the The value of the battery power decrease per unit time to determine whether another power warning is issued.

Such as the above-mentioned computer program product, in which, for the training of the battery power prediction model, the computer program product is loaded into the program by the computer to execute: a power consumption detection command to enable the processor to detect the portable device A total power consumption value per unit time when the one or more application programs are running; a weight adjustment instruction causes the processor to adjust the equivalent power consumption weight of the one or more application programs, so that the battery power prediction model is based on the one to The battery power reduction value per unit time calculated by the equivalent power consumption weights of multiple applications can be close to the total power consumption value per unit time; wherein, the processor repeatedly executes the weight adjustment instruction until the unit time battery The difference between the value of battery power decline and the total power consumption per unit time is less than or equal to a threshold; a weight setting command, when the difference between the battery power decline value per unit time and the total power consumption per unit time is less than or equal to the threshold, The processor sets and stores the current equivalent power consumption weight as the equivalent power consumption weight corresponding to the one or more application programs.

Such as the above-mentioned computer program product, wherein the battery warning is a display screen of the portable device to display a message that a power cord needs to be connected.

Such as the above-mentioned computer program product, wherein the other power warning is a display screen of the portable device to display a message that a power cord needs to be carried.

Such as the above-mentioned computer program product, wherein the decrease in battery power per unit time is the power consumption of the battery per hour, or the percentage of power decrease per hour of the battery.

The method for predicting battery power disclosed in the present invention is applicable to a portable device. The portable device may include: personal digital assistants (PDA), smart phones, laptops, and tablet computers. (tablet)...etc. Figure 1 is a flowchart of the method for predicting battery power according to an embodiment of the disclosure. The portable device uses a battery power prediction model contained in it to determine whether to issue a power warning. Figure 1 is a flowchart of the method for predicting battery power according to an embodiment of the disclosure. As shown in Figure 1, first, the method for predicting battery power of the present invention detects one or more applications currently running on the portable device, and obtains the names of the one or more applications (step S100). For example, the battery power prediction method of the present invention can detect the programs opened by the user on the portable device, the process tree formed, the hardware settings (for example, BIOS settings), and The user's specific program schedule... etc., and at the same time obtain the name of the program opened by the user on the portable device. For example, when the battery power prediction method of the present invention detects that the program being executed by the user is program A, program B, and program C, the battery power prediction method of the present invention also obtains the list of programs being executed as programs A, program B, and program C.

Next, the battery power prediction method of the present invention inputs the names of the one or more applications being executed into the battery power prediction model to calculate a battery power decline value per unit time (step 102). It is worth noting that the battery power prediction model is trained to store equivalent power consumption weights corresponding to the one or more applications. FIG. 2 is a schematic diagram of the battery power prediction model storing equivalent power consumption weights in the embodiment of the disclosure. As shown in Figure 2, the battery power prediction model disclosed in the present invention stores data a, data b, data c, and data d, but the invention is not limited to this. Among them, data a records that when the portable device is running program A, the battery power prediction model sets the equivalent power consumption weight (W _A ) of program A to 1, that is, W _A =1. Data b records that when the portable device executes program A and program B at the same time, the battery power prediction model sets the equivalent power consumption weight (W _A ) of program _A to 0.6 and sets the equivalent power consumption weight of program B (W _B ) is set to 0.4, that is, W _A =0.6 and W _B =0.4. In other words, when the portable device is running both program A and program B, the power consumption of program A accounts for 60% of the current total power consumption of the portable device, and the power consumption of program B accounts for This represents 40% of the current total power consumption of the portable device.

Similarly, data c records that when the portable device executes program A, program B, and program C at the same time, the battery power prediction model sets the equivalent power consumption weight (W _A ) of program _A to 0.4, and program B The equivalent power consumption weight (W _B ) of is set to 0.4, and the equivalent power consumption weight (W _C ) of formula C is set to 0.2, that is, W _A =0.4, W _B =0.4, and W _C =0.2. Data d records that when the portable device executes program A, program B, program C, and program D at the same time, the battery power prediction model sets the equivalent power consumption weight (W _A ) of program _A to 0.3 and program B Set the equivalent power consumption weight (W _B ) of formula C to 0.2, set the equivalent power consumption weight (W _C ) of formula _C to 0.4, and set the equivalent power consumption weight (W _D ) of formula D to 0.1, also That is, W _A =0.3, W _B =0.2, W _C =0.4, W _D =0.1. In some embodiments, the battery power prediction model of the battery power prediction method of the present invention correspondingly finds the equivalent power consumption weight of the one or more application programs according to the difference of the input name of the one or more application programs .

For example, as shown in Figure 2, when the portable device is running program A, program B, and program C, the battery power prediction method of the present invention will be the name of program A, program B, and program C Input to the battery power prediction model, so that the battery power prediction model can correspondingly find the equivalent power consumption weights set when the program A, the program B, and the program C are executed at the same time, that is, W _A =0.4, W _B =0.4 、W _C =0.2, and the battery power prediction model can calculate the battery power decrease value per unit time when the portable device is executing program A, program B, and program C based on the above-mentioned equivalent power consumption weight. The above-mentioned equivalent power consumption weight value is only an example, and is not a limitation of the present invention. In some embodiments, the decrease value of the battery power per unit time may be the power consumption of the battery of the portable device per hour, or the percentage of power decrease of the battery per hour.

Finally, the battery power prediction method of the present invention judges whether to issue the power warning according to the battery power decrease value per unit time (step S104). For example, the battery power prediction model calculates the current battery power decline value per unit time based on the respective equivalent power consumption weights of the one or more applications, and based on the battery power decline value per unit time, Calculate the remaining battery power of the portable device. When the remaining battery power of the portable device is lower than a threshold, the battery power prediction method of the present invention issues a power warning to the user. For example, the battery power prediction method of the present invention can display a message "need to connect a power cord" through a display screen of the portable device.

FIG. 3 is a schematic diagram of the equivalent power consumption weight of the embodiment of the disclosure. As shown in Figure 3, the example (1) is to input the name of program A into the battery power model 300, so that the battery power model 300 is based on the equivalent power consumption weight W _{A of the} program A (that is, the power consumption when the program A is executed alone) Quantity, W _A =1) Calculate the value 302 of the decrease in battery power per unit time. Example (2) is to input the names of program A and program B into the battery power model 300, so that the battery power model 300 is based on the equivalent power consumption weight W _A of the program _A and the equivalent power consumption weight W _{B of the} program _B (simultaneous execution Formula A and Formula B, for example, WA=0.6, WB=0.4) calculate the battery power decrease value 302 per unit time.

Example (3) is to input the names of program A, program B, and program C into the battery power model 300. Different from the above examples (1) and (2), the power consumption network constructed by the example (3) has three layers. In other words, the battery power prediction model 300 needs to first calculate the power consumption value of node A'per unit time according to the power consumption weights of program A, program B, and program C (for example, power consumption weights 304, 306, 308) , And calculate the power consumption value of node B'per unit time according to the power consumption weight of formula C (for example, power consumption weight 310). Finally, the battery power model 300 calculates the magnitude of the battery power decrease per unit time according to the power consumption weight of node A'(for example, power consumption weight 312) and the power consumption weight of node B'(for example, power consumption weight 314) 302. In some embodiments, the battery power model 300 of the present invention can simplify the calculation of Example (3) to be based on the equivalent power consumption weight W _A of the formula _A , the equivalent power consumption weight W _B of the formula _B , and the formula C The equivalent power consumption weight WC is used to calculate the decrease range value 302 of the battery power per unit time. In other words, the battery power prediction method of the present invention takes the power consumption weight 304, the power consumption weight 306, the power consumption weight 308, the power consumption weight 310, the power consumption weight 312, and the power consumption weight 314 to be all equivalent to the formula A. The equivalent power consumption weight W _A contributed, the equivalent power consumption weight W _B contributed by the formula _B , and the equivalent power consumption weight W _C contributed by the formula C are used to reduce the storage in the battery power prediction model 300 The amount of data for the equivalent power consumption weight.

Similarly, example (4) is to input the names of program A, program B, program C, and program D into the battery power model 300. Similar to the above example (3), the power consumption network constituted by the example (4) is also three layers. In other words, the battery power prediction model 300 needs to first calculate the power consumption per unit time of node A'according to the power consumption weights of program A, program B, program C, or program D (for example, power consumption weights 316, 318) Calculate the power consumption value of node B'per unit time according to the power consumption weights of formula B, formula C and formula D (for example, power consumption weights 320, 322, 324), and according to the power consumption weight of formula D ( For example, the power consumption weight 326) is used to calculate the power consumption value of node C'per unit time. Finally, the battery power model 300 is based on the power consumption weight of node A'(for example, power consumption weight 328), the power consumption weight of node B'(for example, power consumption weight 330), and the power consumption weight of node C'(for example, , The power consumption weight 332) to calculate the battery power decrease value 302 per unit time. In some embodiments, as shown in FIG. 3, the number of applications input to the battery power prediction model 300 and the number of corresponding power consumption network layers are only examples, and not a limitation of the present invention.

FIG. 4 is another flowchart of the method for predicting battery power according to an embodiment of the disclosure. The battery power prediction method of the present invention further detects the current time point to determine whether the current time point falls within an early warning period before a scheduled event occurs (step S400). At the moment when the scheduled event occurs, the portable device will execute the one or more applications. For example, according to the schedule, the user will have a meeting at 03:00 PM every Wednesday, and the portable device will execute program A, program B, and program C at the same time during the meeting. Then, when the current time point falls within the warning period before the scheduled event occurs, the battery power prediction method of the present invention inputs the names of the one or more applications into the battery power prediction model to predict the The battery power decrease value 302 per unit time when the scheduled event occurs (step S402). For example, the battery power prediction method of the present invention can determine that the current time point has fallen within the early warning period before the scheduled event (ie meeting) occurs at 02:50 PM on Wednesday (that is, the early warning period is 10 minutes), then The battery power prediction method of the present invention can input the names of program A, program B, and program C into the battery power model 300 at 02:50PM on Wednesday. Since the battery power model 300 has stored the equivalent power consumption weights corresponding to each program when executing program A, program B, and program C at the same time, it can be predicted that users will execute program A, program B, After the program C, the battery power decrease value per unit time of the portable device.

Finally, the method for predicting battery power of the present invention determines whether another power warning is issued according to the value 302 of the battery power decrease per unit time (step S404). In some embodiments, the battery power prediction method of the present invention can display a message "need to carry a power cord" through a display screen of the portable device.

FIG. 5 is a flowchart of the training method of the battery power prediction model according to the disclosed embodiment. As shown in Figure 5, the training method of the battery power prediction model 300 of the present invention first detects a total power consumption per unit time (P _total ) when the one or more applications are running on the portable device (step S500). For example, the battery power prediction model 300 of the present invention can obtain the total power consumption value per unit time of the one or more applications currently running through a third-party power management application. In other words, the total power consumption value per unit time can be regarded as an actual power consumption measurement value obtained by the battery power prediction model 300 of the present invention. Then, the training method of the battery power prediction model 300 of the present invention adjusts the equivalent power consumption weights of the one or more applications, so that the battery power prediction model 300 calculates the equivalent power consumption weights of the one or more applications The battery power decrease range value 302 per unit time can be close to the total power consumption value per unit time (P _total ) (step S502). For example, when the battery power prediction model 300 is training, the battery power prediction model 300 detects that the portable device is executing program A, program B, program C, and program D at the same time. At this time, battery performing predictive model 300 that has a separate power consumption P _A program A, program B execution power consumption P _B alone, the program execution of the power consumption P C _C alone, and a separate execution program D Power consumption P _D. The battery power prediction model 300 individually adjusts the equivalent power consumption weight W _A of formula _A , the equivalent power consumption weight W _B of formula _B , the equivalent power consumption weight W _C of formula _C , and the equivalent power consumption weight of formula D W _D , so that the following formula is established, P _A *W _A +P _B *W _B +P _C *W _C +P _D *W _D ≒P _total .

Furthermore, the training method of the battery power prediction model 300 of the present invention repeatedly adjusts the equivalent power consumption weights of one to a plurality of applications until the difference between the unit time battery power decrease value 302 and the unit time total power consumption value It is less than or equal to a threshold value (P _th ) (step S504). And, when the difference between the battery power decrease value per unit time 302 and the total power consumption value per unit time (P _total ) is less than or equal to the threshold (P _th ), then the current equivalent power consumption weight is set and stored to correspond to The equivalent power consumption weight of the one or more applications. For example, when |P _total- (P _A *W _A +P _B *W _B +P _C *W _C +P _D *W _D )|≦P _th , the battery power model 300 of the present invention will be W _A Set the equivalent power consumption weight of program A in the case of executing program A, program B, program C, and program D at the same time, and set W _B to the case of executing program A, program B, program C, and program D at the same time The equivalent power consumption weight of program B below, set W _C to the equivalent power consumption weight of program C when program A, program B, program C, and program D are executed at the same time, and set W _D to simultaneously The equivalent power consumption weight of program D when executing program A, program B, program C, and program D, and store the above-mentioned W _A , W _B , W _C , and W _D values in the portable device In the memory, for example, data d in Figure 2. Therefore, when the battery power prediction method of the present invention executes step S100 and step S102 in Figure 1, or steps S400 and S402 in Figure 4, the battery power prediction model 300 of the present invention can be based on the detected Find the corresponding equivalent power consumption weights W _A , W _B , W _C , W _D for the names of program A, program B, program C, and program D to calculate the battery power decrease value per unit time 302.

The present invention further discloses a computer program product, which is suitable for a portable device to determine whether a power warning is issued. The computer program product is loaded into the program by a computer to execute: a first detection command, a call command, A first input instruction, and a first judgment instruction. The first detection instruction causes a processor of the computer to execute step S100 in FIG. 1. The call instruction causes the processor to obtain a battery power prediction model from the computer's storage. Wherein, the battery power prediction model is trained to store equivalent power consumption weights corresponding to the one or more applications, and the battery power prediction model finds correspondingly according to the input name of the one or more applications. The equivalent power consumption weights of the one or more applications are calculated, as shown in Figures 2 and 3. The first input instruction causes the processor to execute step S102 in FIG. 1. The first judgment instruction causes the processor to execute step S104 in FIG. 1.

According to the computer program product disclosed in the present invention, it further executes: a second detection command, a second input command, and a second judgment command. The second detection instruction causes the processor to execute step S400 in FIG. 4. The second input instruction causes the processor to execute step S402 in FIG. 4. The second judgment instruction causes the processor to execute step S404 in FIG. 4.

According to the computer program product disclosed in the present invention, for the training of the battery power prediction model, the computer program product loads the program through the computer to execute: a power consumption detection command, a weight adjustment command, and a weight setting command. The power consumption detection instruction causes the processor to execute step S500 in FIG. 5. The weight adjustment instruction causes the processor to execute steps S502 and S504 in FIG. 5. The weight setting instruction causes the processor to execute step S506 in FIG. 5. The steps or actions performed by the computer program product of the present invention can directly correspond to the battery power prediction method disclosed in the present invention. The battery power prediction method disclosed in the present invention is described in paragraphs [0017]~[0028] of the specification of this application. Do a detailed description, so I won't repeat it.

The expected utility of the battery power prediction method and its computer program products disclosed in the present invention are as follows: 1. When the user’s electronic device (for example, a portable device) is not connected to a power source, the user When opening any program, you can infer whether to connect the power cord based on the trained battery power prediction model; 2. Regardless of whether the electronic device is connected to the power cord, when the user has a certain schedule (for example: meeting, business trip, Or class... etc.) When the start time is approaching, users can be advised whether it is necessary to bring a power cord.

Although the embodiments of the present invention are as described above, we should understand that what is presented above is only an example, not a limitation. According to this embodiment, many changes of the above exemplary embodiment can be implemented without violating the spirit and scope of the invention. Therefore, the breadth and scope of the present invention should not be limited by the embodiments described above. More precisely, the scope of the present invention should be defined by the following patented scope and its equivalents.

S100, S102, S104: steps a, b, c, d: data A, B, C, D: program W _A , W _B , W _C , W _D : equivalent power consumption weight 300: battery power prediction model 302: Battery power decrease value per unit time 304, 306, 308, 310: power consumption weight 312, 314, 316, 318: power consumption weight 320, 322, 324, 326: power consumption weight 328, 330, 332: power consumption weight A ', B', C': nodes (1), (2), (3): example S400, S402, S404: steps S500, S502, S504, S506: steps

Figure 1 is a flowchart of the method for predicting battery power according to an embodiment of the disclosure. FIG. 2 is a schematic diagram of the battery power prediction model storing equivalent power consumption weights in the embodiment of the disclosure. FIG. 3 is a schematic diagram of the equivalent power consumption weight of the embodiment of the disclosure. FIG. 4 is another flowchart of the method for predicting battery power according to an embodiment of the disclosure. FIG. 5 is a flowchart of the training method of the battery power prediction model according to the disclosed embodiment.

S100, S102, S104: steps

Claims (10)

A method for predicting battery power is suitable for a portable device. The portable device uses a battery power prediction model contained in it to determine whether to issue a power warning. The prediction method includes: detecting the battery power One or more applications currently running on the portable device, and obtain the names of the one or more applications; input the names of the one or more applications into the battery power prediction model to calculate the battery power per unit time Decline value; wherein, the battery power prediction model is trained to store equivalent power consumption weights corresponding to the one or more applications; the battery power prediction model is based on the input name of the one or more applications, and Correspondingly find out the equivalent power consumption weights of the one or more applications; determine whether the battery power warning is issued according to the value of the battery power drop per unit time; the prediction method further includes: detecting the current time point to determine Whether the current time falls within an early warning period before a scheduled event occurs; where the portable device will execute the one or more applications at the moment when the scheduled event occurs; when the current time has When it falls within the warning period before the scheduled event, input the names of the one or more applications into the battery power prediction model to predict the battery power decline value per unit time when the scheduled event occurs; according to the unit The value of the time the battery power drops to determine whether to send another power Quantity warning. Such as the prediction method described in claim 1, wherein the training method of the battery power prediction model includes: detecting the total power consumption per unit time when the one or more applications are running on the portable device Value; adjust the equivalent power consumption weights of the one or more applications, so that the battery power prediction model calculated based on the equivalent power consumption weights of the one or more applications can be similar to The total power consumption value per unit time; repeatedly adjust the equivalent power consumption weights of one to multiple applications until the difference between the battery power decrease value per unit time and the total power consumption value per unit time is less than or equal to a threshold; If the difference between the battery power decrease value per unit time and the total power consumption value per unit time is less than or equal to the threshold, the current equivalent power consumption weight is set and stored as the equivalent power consumption weight corresponding to the one or more applications. For example, in the prediction method described in item 1 of the scope of patent application, the power warning is a display screen of the portable device to display a message that a power cord needs to be connected. For example, in the prediction method described in item 1 of the scope of patent application, the other battery warning is displayed through a display screen of the portable device, indicating that a power cord needs to be carried. For example, the prediction method described in item 1 of the scope of patent application, wherein the decrease in battery power per unit time is the power consumption of the battery per hour, or the battery The percentage of electricity dropped per hour. A computer program product suitable for a portable device to determine whether to issue a power warning. The computer program product is loaded into the program by a computer to execute: a first detection command to enable a processor of the computer to detect Measure one or more applications currently running on the portable device, and obtain the name of the one or more applications; a call command enables the processor to obtain a battery power prediction model from the computer's memory; Wherein, the battery power prediction model is trained to store the equivalent power consumption weights corresponding to the one or more applications; the battery power prediction model finds out correspondingly according to the input names of the one or more applications The equivalent power consumption weight of the one or more application programs; a first input command causes the processor to input the name of the one or more application programs into the battery power prediction model to calculate the battery power decrease per unit time Value; a first judgment command that enables the processor to determine whether to issue the power warning according to the value of the battery power drop per unit time; the computer program product execution further includes: a second detection command to make the processor detect The current time point is used to determine whether the current time point falls within an early warning period before the occurrence of a scheduled event; wherein, at the moment when the scheduled event occurs, the portable device will execute the one or more applications ; A second input instruction, when the current time point falls into the scheduled event During the previous warning period, the processor is made to input the names of the one or more applications into the battery power prediction model to predict the battery power decline value per unit time when the scheduled event occurs; a second judgment instruction , Enabling the processor to determine whether to issue another battery power warning based on the battery power decrease value per unit time. For example, the computer program product described in item 6 of the scope of patent application, in which, for the training of the battery power prediction model, the computer program product is loaded into the program by the computer to execute: a power consumption detection command to make the processor detect Measure the total power consumption per unit time when the one or more application programs are running on the portable device; a weight adjustment command causes the processor to adjust the equivalent power consumption weight of the one or more application programs so that the The battery power prediction model calculated based on the equivalent power consumption weights of the one or more applications, the battery power decline value per unit time can be close to the total power consumption per unit time; wherein the processor repeatedly executes the weight adjustment Command until the difference between the battery power decrease value per unit time and the total power consumption value per unit time is less than or equal to a threshold; a weight setting command, when the battery power decrease value per unit time and the total power consumption value per unit time The difference is less than or equal to the threshold, so that the processor sets and stores the current equivalent power consumption weight as the equivalent power consumption weight corresponding to the one or more application programs. For example, the computer program product described in item 6 of the scope of patent application, wherein the battery warning is displayed through a display screen of the portable device, indicating that a power supply needs to be connected Line message. For example, the computer program product described in item 6 of the scope of patent application, wherein the other battery warning is a display screen of the portable device to display a message that a power cord needs to be carried. For example, the computer program product described in item 6 of the scope of patent application, wherein the decrease in battery power per unit time is the power consumption of the battery per hour, or the percentage of power decrease per hour of the battery.

Images