CN-121997001-A - Model parameter updating method and device based on vehicle network interaction and electronic equipment

CN121997001ACN 121997001 ACN121997001 ACN 121997001ACN-121997001-A

Abstract

The invention discloses a model parameter updating method and device based on vehicle network interaction and electronic equipment. The method comprises the steps of obtaining characteristic characterization parameters corresponding to a plurality of target terminals respectively, determining prediction output of a power scheduling model according to the characteristic characterization parameters corresponding to the plurality of target terminals respectively, determining model output deviation according to the prediction output and reference output of the power scheduling model, determining update demand parameters according to the model output deviation, dividing the update demand parameters into a plurality of sub-demand parameters to obtain update operation parameters corresponding to the sub-models corresponding to the plurality of target terminals respectively, and updating the model parameters of the power scheduling model according to the update operation parameters of the sub-models corresponding to the plurality of target terminals respectively. The method solves the technical problem that in the related art, when model parameters are updated in a vehicle network interaction scene, the data of the multiparty parameter main body are required to be acquired for collaborative analysis, so that the data leakage exists.

Inventors

JIN YUAN
MA KAI
WANG WEI
GUAN YU
GU JUN
LI ZHUOQUN
SUN QINFEI
ZHANG QIAN
LIU XIANGLU
YUAN XIAOXI
WAN YI
WANG HONGBIAO

Assignees

国网北京市电力公司

Dates

Publication Date: 20260508
Application Date: 20260121

Claims (10)

1. The model parameter updating method based on the vehicle network interaction is characterized by comprising the following steps of: obtaining characteristic characterization parameters corresponding to a plurality of target terminals respectively in a vehicle network interaction scene, wherein the corresponding characteristic characterization parameters are used for characterizing terminal characteristics of the corresponding target terminals; Determining the prediction output of the power scheduling model according to the characteristic characterization parameters respectively corresponding to the target terminals; Determining a model output deviation of the power scheduling model according to the prediction output and the reference output of the power scheduling model, wherein the power scheduling model comprises a plurality of sub-models respectively corresponding to a plurality of target terminals; Determining an update demand parameter corresponding to the power dispatching model according to the model output deviation, wherein the update demand parameter is used for representing the demand characteristic of the power dispatching model for model parameter update; Dividing the update demand parameters into a plurality of sub demand parameters, respectively issuing the sub demand parameters to corresponding target terminals to obtain update operation parameters corresponding to sub models respectively corresponding to the plurality of target terminals, wherein the corresponding update operation parameters are used for representing operation characteristics of model parameter update on the sub models of the corresponding target terminals; And updating the model parameters of the power scheduling model according to the updating operation parameters of the sub-models respectively corresponding to the target terminals.
2. The method of claim 1, wherein before the obtaining the characteristic characterization parameters corresponding to the plurality of target terminals in the vehicle-network interaction scenario, the method comprises: determining a power grid terminal and a plurality of equipment terminals in the vehicle-network interaction scene; determining matching indexes among the plurality of equipment terminals according to encryption identification parameters respectively corresponding to the plurality of equipment terminals; Determining a plurality of terminal groups from the plurality of equipment terminals according to the matching indexes among the plurality of equipment terminals, wherein the plurality of terminal groups respectively comprise a first terminal corresponding to the electric automobile and a second terminal corresponding to the regional equipment; And determining the target terminals according to the terminal groups and the power grid terminals.
3. The method according to claim 2, wherein before determining the matching indexes among the plurality of device terminals according to the encrypted identification parameters corresponding to the plurality of device terminals, the method comprises: determining time stamps and position coordinates corresponding to the plurality of equipment terminals respectively; determining space-time identification parameters corresponding to the plurality of equipment terminals according to the time stamps and the position coordinates corresponding to the plurality of equipment terminals respectively, wherein the corresponding space-time identification parameters are identification parameters for representing space-time characteristics of the corresponding equipment terminals; And encrypting the space-time identification parameters respectively corresponding to the plurality of equipment terminals to obtain encrypted identification parameters respectively corresponding to the plurality of equipment terminals.
4. The method according to claim 1, wherein determining the predicted output of the power scheduling model according to the characteristic characterization parameters corresponding to the plurality of target terminals, respectively, comprises: Under the condition that the plurality of target terminals comprise a power grid terminal, a first terminal corresponding to the electric automobile and a second terminal corresponding to the regional equipment, determining a first encryption output based on the characteristic characterization parameters corresponding to the first terminal; Determining a second encryption output based on the first encryption output and the characteristic characterization parameters corresponding to the second terminal; and determining the prediction output of the power dispatching model based on the second encryption output and the characteristic characterization parameters corresponding to the power grid terminal.
5. The method according to claim 1, wherein the updating the model parameters of the power scheduling model according to the updated operation parameters of the sub-models respectively corresponding to the plurality of target terminals includes: determining a parameter updating amount corresponding to the power scheduling model according to updating operation parameters respectively corresponding to the plurality of target terminals; and updating the model parameters of the power dispatching model according to the parameter updating quantity.
6. The method according to claim 5, wherein after updating the model parameters of the power scheduling model according to the updated operation parameters of the sub-models respectively corresponding to the plurality of target terminals, the method further comprises: updating the model parameters of the power dispatching model to obtain updated parameters, wherein the updated parameters comprise a plurality of sub-updated parameters; and respectively issuing the plurality of sub-updating parameters to corresponding target terminals in the plurality of target terminals to be used for updating the model parameters of the sub-models of the corresponding target terminals.
7. The method according to any one of claims 1 to 6, wherein dividing the update demand parameter into a plurality of sub demand parameters, and respectively issuing the sub demand parameters to the corresponding target terminals to obtain update operation parameters corresponding to sub models respectively corresponding to the plurality of target terminals, includes: Determining forward propagation data of sub-models respectively corresponding to the plurality of target terminals; And carrying out back propagation according to the forward propagation data of the sub-models respectively corresponding to the target terminals and the sub-demand parameters respectively corresponding to the target terminals respectively based on the sub-models respectively corresponding to the target terminals to obtain updated operation parameters corresponding to the sub-models respectively corresponding to the target terminals.
8. Model parameter updating device based on car network interaction, characterized by comprising: The system comprises an acquisition module, a storage module and a storage module, wherein the acquisition module is used for acquiring characteristic characterization parameters corresponding to a plurality of target terminals respectively in a vehicle network interaction scene, wherein the corresponding characteristic characterization parameters are used for characterizing terminal characteristics of the corresponding target terminals; the first determining module is used for determining the prediction output of the power scheduling model according to the characteristic characterization parameters respectively corresponding to the plurality of target terminals; the second determining module is used for determining a model output deviation of the power scheduling model according to the prediction output and the reference output of the power scheduling model, wherein the power scheduling model comprises a plurality of sub-models corresponding to the target terminals respectively; The third determining module is used for determining an update demand parameter corresponding to the power dispatching model according to the model output deviation, wherein the update demand parameter is used for representing the demand characteristic of the power dispatching model for model parameter update; a fourth determining module, configured to divide an update demand parameter into a plurality of sub-demand parameters, and send the sub-demand parameters to corresponding target terminals respectively, so as to obtain update operation parameters corresponding to sub-models respectively corresponding to the plurality of target terminals, where the corresponding update operation parameters are used to characterize operation features of performing model parameter update on the sub-models of the corresponding target terminals; And a fifth determining module, configured to update model parameters of the power scheduling model according to update operation parameters of sub-models corresponding to the target terminals respectively.
9. An electronic device, comprising: A processor; a memory for storing the processor-executable instructions; Wherein the processor is configured to execute the instructions to implement the model parameter updating method based on vehicle network interaction of any one of claims 1 to 7.
10. A computer readable storage medium, characterized in that instructions in the computer readable storage medium, when executed by a processor of an electronic device, enable the electronic device to perform the model parameter updating method based on vehicle network interaction according to any one of claims 1 to 7.

Description

Model parameter updating method and device based on vehicle network interaction and electronic equipment Technical Field The invention relates to the technical field of information security, in particular to a model parameter updating method and device based on vehicle-network interaction and electronic equipment. Background In the related art, in a vehicle-network interaction scene, a power scheduling model is a core tool for realizing power resource scheduling in the vehicle-network interaction scene, and the power scheduling model can be ensured to perform optimal configuration and efficient utilization of power resources based on actual conditions through updating model parameters. However, in the related art, when the model parameter is updated, the data of the multiparty parameter main body needs to be acquired for collaborative analysis, so that the technical problem of data leakage exists. In view of the above problems, no effective solution has been proposed at present. Disclosure of Invention The embodiment of the invention provides a model parameter updating method and device based on vehicle network interaction and electronic equipment, which at least solve the technical problem that in the related art, when model parameters are updated in a vehicle network interaction scene, data of a multiparty parameter main body are required to be acquired for collaborative analysis, so that data leakage exists. According to one aspect of the embodiment of the invention, a model parameter updating method based on vehicle network interaction is provided, which comprises the steps of obtaining characteristic characterization parameters corresponding to a plurality of target terminals respectively in a vehicle network interaction scene, dividing the characteristic characterization parameters into a plurality of sub-demand parameters, respectively transmitting the sub-demand parameters to the corresponding target terminals to obtain update operation parameters corresponding to the sub-models respectively, determining model output deviation of the power dispatching model according to the prediction output and reference output of the power dispatching model, wherein the power dispatching model comprises sub-models corresponding to the target terminals respectively, determining update demand parameters corresponding to the power dispatching model according to the model output deviation, wherein the update demand parameters are used for characterizing demand characteristics of the power dispatching model for model parameter update, dividing the update demand parameters into the sub-demand parameters, respectively transmitting the sub-demand parameters to the corresponding target terminals to obtain update operation parameters corresponding to the sub-models respectively, wherein the corresponding update operation parameters are used for characterizing the sub-models for updating the corresponding target terminals respectively, and updating the sub-model operation parameters of the power dispatching model according to the new sub-model operation parameters. Optionally, before the characteristic characterization parameters corresponding to the plurality of target terminals in the vehicle network interaction scene are obtained, the method comprises the steps of determining a power grid terminal and a plurality of equipment terminals in the vehicle network interaction scene, determining matching indexes among the plurality of equipment terminals according to encryption identification parameters corresponding to the plurality of equipment terminals, determining a plurality of terminal groups from the plurality of equipment terminals according to the matching indexes among the plurality of equipment terminals, wherein the plurality of terminal groups respectively comprise a first terminal corresponding to an electric vehicle and a second terminal corresponding to regional equipment, and determining the plurality of target terminals according to the plurality of terminal groups and the power grid terminal. Optionally, before determining the matching indexes among the plurality of equipment terminals according to the encryption identification parameters respectively corresponding to the plurality of equipment terminals, determining time stamps and position coordinates respectively corresponding to the plurality of equipment terminals, determining space-time identification parameters respectively corresponding to the plurality of equipment terminals according to the time stamps and the position coordinates respectively corresponding to the plurality of equipment terminals, wherein the corresponding space-time identification parameters are identification parameters for representing space-time characteristics of the corresponding equipment terminals, and encrypting the space-time identification parameters respectively corresponding to the plurality of equipment terminals to obtain the encryption identification parameters respe