Automatic parking method based on fuzziness and deep reinforcement learning

Abstract

The invention discloses an automatic parking method based on fuzziness and deep reinforcement learning. The method comprises the following steps of 1, constructing a fuzzy action network, and outputting a control instruction to establish a sample pool set; 2, building a fuzzy evaluation network for training the fuzzy action network; 3, training networks through taking samples [ st, at, Rt and st+1] at the t moment in a sample pool set as input; and 4, assigning a value of t+1 to t to return to the step 3 to continue learning until t is larger than C. According to the automatic parking method,the automatic parking can be completed through a control method of combining a fuzzy neural network and the deep reinforcement learning, so that the automatic parking process is safer and more reliable, and the occurrence of parking accidents is reduced.

Description

technical field [0001] The invention relates to the technical field of automatic parking planning for smart cars, in particular to an automatic parking method based on fuzzy deep reinforcement learning. Background technique [0002] With the continuous increase of the number of motor vehicles, the parking space becomes crowded, and the crowded parking space will bring safety, economy, environment, health and other problems to the city, and the parking problem has become an inevitable problem. At the same time, due to the crowded parking environment and the technical level of drivers, parking accidents occur frequently. With the development of parking technology, car companies have introduced semi-automatic parking systems and fully automatic parking systems. The semi-automatic parking system uses the camera to collect image data and the ultrasonic radar to detect the distance data of the surrounding objects from the vehicle body, and reminds the driver to park through the d...

AI Technical Summary

Problems solved by technology

The semi-automatic parking system uses the camera to collect image data and the ultrasonic radar to detect the distance data of the surrounding objects from the vehicle body, and reminds the driver to park through the data of the sensor, but completes the parking according to the subjective factors of the driver; the automatic parking system adopts the traditional The method of trajectory planning usually includes two-stage or three-stage parking. There are reasons such as sensor misidentification, complex environment, and trajectory error, which cannot be applied to various parking environments.

Images