Control method for inputting "smart acquisition" model into automatic driving
A technology of automatic driving and control method, which is applied in the directions of driver input parameters, non-electric variable control, two-dimensional position/channel control, etc. question
Pending Publication Date: 2020-04-21
顾泽苍
View PDF0 Cites 5 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0006] Difficulties in human-machine sensory fusion in vehicle distance selection: According to a company’s survey, when encountering a self-driving car on the road, 41% of drivers think that the farther away from the self-driving car, the better, but some people think that it is better to follow a fixed distance. The distance is better, or think that it is better to be close, and there are people who are curious to catch up with the self-driving car in front
This is how to solve the problem of human-machine sensory fusion as a self-driving car, and how to choose the driving problem that is closest to human beings, which has become a complex automatic control problem.
[0007] The problem of man-machine rights transfer: In the conversion stage of man-machine operation, the consciousness between man-machine cannot be transferred
[0026] The above (non-patent literature 4) mainly solves the problem of automatic driving of trains. The proposed rule base using fuzzy inference realizes automatic driving of trains. The establishment of a huge knowledge base requires a large-scale manual construction of the rule base. Moreover, it can only solve two or three objective functions, so it is difficult to apply in automatic driving.
[0027] Although the above (non-patent literature 5) proposed multi-purpose fuzzy control, it is also because the fuzzy control used is relatively reluctant to control corresponding to more objective functions, so it still stays at the individual control of each specific objective function
Especially in the fusion of human-machine perception and objective functions such as safety, energy saving, and comfort in self-driving cars, and the simultaneous control of multiple objectives, since different objective functions are not in the same space, it is impossible to find a common optimal control point , even if they are mapped to the same space, it is impossible to obtain a common optimal intersection point for different objective functions using traditional methods. Therefore, it is necessary to find the redundancy between multi-purpose optimal controls, and truly achieve multi-purpose optimal control. Therefore , the establishment of a machine learning model that needs to address multi-purpose control
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0090] The embodiments of the present invention will be further described below in conjunction with the accompanying drawings, but the embodiments of the present invention are illustrative rather than limiting.
[0091] figure 1 It is a schematic diagram of the approximation method of road lanes realized by automatic machine learning.
[0092] The SDL (Super Deep Learning) model proposed in this application refers to a self-organizing machine learning model composed of a plurality of probability scales, or a plurality of automatic machine learning models, and uses a distance formula that can unify Euclidean space and probability space, Or an artificial intelligence system composed of all or part of the models in the fuzzy event probability measurement formula that can unify Euclidean space and probability space.
[0093] Such as figure 1 As shown in (a): This is an unsupervised machine learning model similar to the definition of self-organization of the probability scale a...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The invention provides a control method for inputting a "smart acquisition" model into automatic driving. The method is characterized in that a state instruction of a relationship with a passing vehicle or a road condition is obtained through consciousness determination; after the state instruction is obtained, "smart acquisition" data corresponding to the state are invoked, and automatic drivingvehicle is controlled to run according to the "smart acquisition" data. The method has the effects that the NP problem caused by extremely high complexity of automatic driving control at present can be solved; a contradiction problem of safe driving and comfortable riding in the field of automatic driving at present can be solved, and a problem that automatic driving accidents continuously occur because safety is sacrificed to pursue comfort of automatic driving at present can be solved. And the effect that "smart acquisition" and "consciousness determination" are integrated with "comfortableriding" and "fast arrival" can be achieved.
Description
【Technical field】 [0001] The invention belongs to a control method for introducing a "witty acquisition" model into an automatic driving in the field of artificial intelligence. 【Background technique】 [0002] Autonomous driving is the main battlefield of artificial intelligence, but unfortunately there are very few research results on dedicated machine learning for autonomous driving applications, and it has not attracted widespread attention so far. [0003] The well-known Japanese company Toyota has published a patent (Patent Document 1) on the "Driving Direction Estimation Device". This patent proposes that according to the automatic driving process of the car, in response to unexpected situations, even if the driver does not respond, the artificial intelligence The machine learning algorithm of inverse transfer neural network automatically selects the driving state to avoid driving accidents, etc. [0004] On October 9, 2016, Japan's NHK commentary committee member Tet...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More Patent Type & AuthorityApplications(China)
IPC IPC(8): G05D1/02G06N99/00
CPCG05D1/0221G06N20/00G06N7/023B60W60/001B60W2556/10B60W2540/12B60W2540/10B60W2540/18B60W40/09G06N7/01G05D1/0088G06N5/046
Inventor顾泽苍
Owner顾泽苍