CN-122026587-A - Power supply control method, device, equipment and medium

CN122026587ACN 122026587 ACN122026587 ACN 122026587ACN-122026587-A

Abstract

The application relates to the field of power supply systems, in particular to a power supply control method, a device, equipment and a medium, and aims to solve the problem that the prior art is difficult to effectively control a complex power supply system. The power supply system control method comprises the steps of obtaining power supply state information, inputting the power supply state information into a preset multi-agent model to obtain power supply control information, and controlling and processing a power supply system according to the power supply control information.

Inventors

YIN JINGYUAN
WEI MENG
WU LIXIN
YANG JIE
HUO QUNHAI
WEI TONGZHEN

Assignees

中国科学院电工研究所

Dates

Publication Date: 20260512
Application Date: 20241107

Claims (10)

1. A power supply control method, characterized by comprising: Acquiring power supply state information; inputting the power state information into a preset multi-agent model to obtain power control information; According to the power control information, controlling the power system; Wherein, obtain the multi-agent model of predetermineeing, include: Acquiring current power state information; dividing the current power state information to obtain power state grouping data; Respectively inputting the power state grouping data to corresponding intelligent agents to obtain an action set; Obtaining power state information of the next moment according to the action set; Obtaining a fault state, a bus voltage state, an energy state, a battery pack state and a switch state according to the power state information at the next moment and the current power state information; obtaining a reward value according to the fault state, the bus voltage state, the energy state, the battery pack state and the switch state; Updating the parameters of the intelligent agent according to the rewarding value; after determining that the training round reaches the preset round, storing and processing parameters of a plurality of current intelligent agents to obtain the preset multi-intelligent agent model.
2. The method of claim 1, wherein dividing the current power state information to obtain power state packet data comprises: Dividing the power state information according to data sources to obtain a battery pack data set and other data sets; dividing the other data sets to obtain a piezoelectric power supply data set, a power distribution data set and a cable data set; And combining the battery pack data set, the piezoelectric power supply data set, the power distribution data set and the cable data set to obtain power state grouping data.
3. The method according to claim 1, wherein the obtaining power state information of the next moment according to the action set includes: obtaining the state of charge of the battery unit in the battery pack according to the state of charge of the battery unit in the battery pack at the last moment, the switch duration, the battery capacity and the current of the battery pack; And obtaining the battery voltage according to the battery current, the battery open-circuit voltage, the battery resistance, the first voltage and the second voltage, wherein the first voltage and the second voltage meet a preset relation.
4. The method of claim 3, wherein obtaining the current of the battery pack comprises: And obtaining the current of the battery pack according to the input state, the bus voltage, the load power and the piezoelectric power supply power.
5. The method of claim 1, wherein deriving the prize value based on the fault condition, the bus voltage condition, the energy condition, the battery condition, the switch condition, comprises: Obtaining a first coefficient, a second coefficient, a third coefficient, a fourth coefficient and a fifth coefficient corresponding to the fault state, the bus voltage state, the energy state, the battery state and the switch state respectively according to the fault state, the bus voltage state, the energy state, the battery state and the switch state; and obtaining a reward value according to the fault state, the bus voltage state, the energy state, the battery pack state, the switch state, the first coefficient, the second coefficient, the third coefficient, the fourth coefficient and the fifth coefficient.
6. The method according to claim 1, wherein the obtaining a fault state, a bus voltage state, an energy state, a battery pack state, a switch state according to the power state information at the next time and the current power state information includes: Obtaining a fault state according to the fault state of the piezoelectric power supply, the fault state of the reconfigurable battery pack and the fault state of the power distribution part; Obtaining a bus voltage state according to the bus voltages at two ends of the battery pack and the bus voltages at two ends of the battery pack at the next moment; Obtaining an energy state according to the load power, the piezoelectric power supply power and the battery pack power; obtaining the state of the battery pack according to the state of charge of the battery unit and the average state of charge of the reconfigurable battery pack; and obtaining a switch state according to the current state and the next time state of the switch.
7. The method of claim 1, wherein the power state information comprises: a failure of the piezoelectric power supply, a failure of the battery cell, a failure of the distribution power supply, a failure of the load, a state of charge of the battery cell, a power of the piezoelectric power supply, and a bus voltage of the battery cell.
8. A power supply control apparatus, comprising: The acquisition module is used for acquiring power supply state information; The analysis module is used for inputting the power state information into a preset multi-agent model to obtain power control information; the control module is used for controlling the power supply system according to the power supply control information; Wherein, obtain the multi-agent model of predetermineeing, include: Acquiring current power state information; dividing the current power state information to obtain power state grouping data; Respectively inputting the power state grouping data to corresponding intelligent agents to obtain an action set; Obtaining power state information of the next moment according to the action set; Obtaining a fault state, a bus voltage state, an energy state, a battery pack state and a switch state according to the power state information at the next moment and the current power state information; obtaining a reward value according to the fault state, the bus voltage state, the energy state, the battery pack state and the switch state; Updating the parameters of the intelligent agent according to the rewarding value; after determining that the training round reaches the preset round, storing and processing parameters of a plurality of current intelligent agents to obtain the preset multi-intelligent agent model.
9. A power control apparatus comprising a processor and storage means, the storage means being adapted to store a plurality of program code, characterised in that the program code is adapted to be loaded and executed by the processor to perform the method of any one of claims 1 to 7.
10. A computer readable storage medium, in which a plurality of program codes are stored, characterized in that the program codes are adapted to be loaded and run by a processor to perform the method of any one of claims 1 to 7.

Description

Power supply control method, device, equipment and medium Technical Field The application relates to the field of power supply systems, and particularly provides a power supply control method, a device, equipment and a medium. Background Missiles play a vital role in modern warfare, and their excellent remote precision striking capability makes them critical to mastering battlefield initiatives. With the continued advancement of aviation and aerospace technology, the reliability requirements of missiles are also gradually increasing. The missile comprises a plurality of subsystems such as a warhead, a propeller, guidance control and a power supply. Among the subsystems of the missile, the power supply system is known as the "heart" of the missile, which ensures the normal operation of the subsystems. In the prior art, the control strategy of the missile-borne power supply system is manually formulated on the basis of a large amount of priori knowledge, and the mode is only suitable for the power supply system with a simple structure. As the structure of the power supply system becomes more and more complex, the control strategy of the power supply system is difficult to acquire based on priori knowledge, so that the technical problem that the control strategy obtained through the priori knowledge cannot effectively control the power supply system is caused. Accordingly, there is a need in the art for a new power control scheme to address the above-described problems. Disclosure of Invention The present application has been made to overcome the above-mentioned drawbacks, and aims to provide a solution or at least partially solve the technical problem that the control strategy obtained by prior knowledge cannot effectively control the power supply system in the prior art. In a first aspect, the present application provides a power supply control method, including: Acquiring power supply state information; inputting the power state information into a preset multi-agent model to obtain power control information; According to the power control information, controlling the power system; Wherein, obtain the multi-agent model of predetermineeing, include: Acquiring current power state information; dividing the current power state information to obtain power state grouping data; Respectively inputting the power state grouping data to corresponding intelligent agents to obtain an action set; Obtaining power state information of the next moment according to the action set; Obtaining a fault state, a bus voltage state, an energy state, a battery pack state and a switch state according to the power state information at the next moment and the current power state information; obtaining a reward value according to the fault state, the bus voltage state, the energy state, the battery pack state and the switch state; Updating the parameters of the intelligent agent according to the rewarding value; after determining that the training round reaches the preset round, storing and processing parameters of a plurality of current intelligent agents to obtain the preset multi-intelligent agent model. In one technical scheme of the above power control method, the dividing the current power state information to obtain power state packet data includes: Dividing the power state information according to data sources to obtain a battery pack data set and other data sets; dividing the other data sets to obtain a piezoelectric power supply data set, a power distribution data set and a cable data set; And combining the battery pack data set, the piezoelectric power supply data set, the power distribution data set and the cable data set to obtain power state grouping data. In one technical scheme of the above power control method, the obtaining power state information of the next moment according to the action set includes: obtaining the state of charge of the battery unit in the battery pack according to the state of charge of the battery unit in the battery pack at the last moment, the switch duration, the battery capacity and the current of the battery pack; And obtaining the battery voltage according to the battery current, the battery open-circuit voltage, the battery resistance, the first voltage and the second voltage, wherein the first voltage and the second voltage meet a preset relation. In one aspect of the above power control method, obtaining the current of the battery pack includes: And obtaining the current of the battery pack according to the input state, the bus voltage, the load power and the piezoelectric power supply power. In one technical scheme of the above power supply control method, the obtaining the reward value according to the fault state, the bus voltage state, the energy state, the battery pack state, and the switch state includes: Obtaining a first coefficient, a second coefficient, a third coefficient, a fourth coefficient and a fifth coefficient corresponding to the fault state, the bus voltage state, the e