Dynamic power supply management method

Abstract

The invention discloses a dynamic power supply management method. An on-line estimation technology is adopted, a moving track of a controlled object is recorded, gradient information of system power consumption and performance about management strategy parameters is learned and estimated in real time, and a management strategy is adaptively regulated through the gradient information, so that the power supply management strategy is adaptively regulated in real time based on the moving track of a system, and personalized dynamic power supply management is realized finally.

Description

Technical field: [0001] The invention relates to a dynamic power management method. Background technique: [0002] In daily life, electronic devices such as mobile phones, laptops, and computers have become indispensable daily necessities for people. A large number of electronic products consume a lot of energy every day, reducing the power consumption of electronic equipment will effectively save energy. Although the individual power consumption of these electronic devices is not very large, the energy consumed by a large number of electronic devices is indeed a very large number. Effectively reducing the energy consumption of each individual will save a lot of energy. In addition, this type of electronic equipment mostly uses batteries as energy storage components, and the storage capacity of conventional batteries has basically reached its limit. Reducing the power consumption of equipment will help prolong the use time of the same power, and bring convenience to people...

AI Technical Summary

Problems solved by technology

The switching time in the overtime technology is artificially set based on experience. If the time is set too small, it will cause frequent switching and more power consumption; if the time is set too large, it will not be able to effectively save energy and consume a lot of energy. on idle

In order to improve the shortcomings of timeout technology, predictive management technology has been applied. The mechanism is that when the power manager uses historical data to predict that the device may be idle for a long time, the power management strategy forces the system to enter a low power consumption state. The method usually uses a nonlinear decay equation to predict the length of the next idle time, once the wrong prediction may lead to system performance degradation and battery waste

Time-out and predictive strategies are empirical strategies, so they cannot achieve the best effect of energy saving

In recent years, the dynamic power management technology based on stochastic model has received extensive attention. It can not only guarantee the optimal energy saving effect, but also ensure that the service quality of the equipment is not greatly affected. However, the existing schemes need to accurately establish mathematical Random model, if the model is not established accurately, the dynamic power management strategy cannot effectively save energy

Images