In-vehicle interactive control algorithm based on deep learning

Abstract

The invention relates to a vehicle-mounted interaction control algorithm based on deep learning, and belongs to the technical field of intelligent control of the internet of vehicles. The algorithm includes the steps of sparsely representing natural language data under multi-feature fusion so that word vectors, part of speech, dependence relationship and other context vectors in words, sentences and other multi-layer structures are deeply fused, and sparsely representing fused word vectors; optimizing a natural language deep learning model under half-monitored learning so that the deep learning model can be matched with a natural language task, and rapidly switching a gradient learning mechanism according to different training environments; establishing and deducing a natural language tensor knowledge map under representation learning so that the entity, concept, classification and semantic relationship in the knowledge map can be converted into numerical vectors in the same semantic space, obtaining knowledge feature vectors through multi-column convolution neural network learning, and conducting knowledge deducing through calculation of the similarity score of target question vectors and the knowledge feature vectors. The natural language information of user interaction can be effectively responded to, and the driving requirements of a user are met.

Description

technical field [0001] The invention belongs to the technical field of intelligent control of the Internet of Vehicles, and relates to a vehicle-mounted interactive control algorithm based on deep learning. Background technique [0002] With the rapid development of the automobile industry, automobiles continue to penetrate into people's lives. At the same time, as in-vehicle smart devices provide drivers with a lot of convenience and comfortable experience, consumers' demand for in-vehicle devices is increasing, and their performance requirements are gradually increasing. [0003] In the prior art, the system operated and controlled by buttons or keyboards can only realize basic control functions, and repeated operations are required for SMS and navigation, which is not conducive to driving safety and brings poor user experience to users. However, the current use of voice recognition control mostly uses a method similar to character string matching for user interactive con...

AI Technical Summary

Problems solved by technology

[0003] In the prior art, the system operated and controlled by buttons or keyboards can only realize basic control functions, and repeated operations are required for SMS and navigation, which is not conducive to driving safety and brings poor user experience to users

However, the current use of voice recognition control mostly uses a method similar to character string matching for user interactive control systems. Drivers need to memorize and accurately call specific voice information to realize the control function. The lack of humanized human-computer interaction design has a negative impact on the driving experience. Complicated and cumbersome, which reduces user experience and is not conducive to driving safety

Images