KẾT HỢP HỌC SÂU VÀ HỌC TĂNG CƯỜNG ỨNG DỤNG TRONG LÁI XE TỰ ĐỘNG

0 Người đánh giá. Xếp hạng trung bình 0

Tác giả: Van Chinh Bui, Thi Luyen Cao, Minh Huy Nguyen, Quang Duc Nguyen

Ngôn ngữ: vie

Ký hiệu phân loại:

Thông tin xuất bản: Tạp chí Khoa học & Công nghệ - Trường Đại học Công nghiệp Hà Nội, 2022

Mô tả vật lý: tr.68

Bộ sưu tập: Metadata

ID: 337464

Thêm vào giỏ Liên kết toàn văn

Convolutional Neural Network (CNN) can detect images from cameras installed on self-driving cars.First, we drove a car on a simulator and recorded frames from three cameras: left, right, and center.These frames were recorded at the rate of 30 frames per second. Additional data recorded were thedistribution of steering angles, average velocity, etc. were passed through a CNN to train a self-drivingsystem. CNN is like the eyes and visual area in the brain, so CNN's achievements in autonomous vehiclecontrol are somewhat limited. Therefore, this paper proposes the use of algorithms based on DeepLearning (DL) combined with reinforcement learning (RL) in the control of autonomous vehicles. Wecall this algorithm Deep Reinforcement Learning (DRL) which can send control commands to the vehicleto navigate properly and efficiently along a defined route. CNN tracks multiple objects while RL predictsthe environment or assesses the current condition of the vehicle to make the safest decision. DRL-basedalgorithms have been used to solve Markov Decision Processes (MDPs), where the scope of thealgorithm is to compute the optimal policy of an autonomous vehicle for choosing actions in aenvironment with the goal of maximizing a reward function.Trong bài báo này, chúng tôi nghiên cứu cách ứng dụng mạng nơ ron tích chập (Convolutionalneural networks - CNN) trong việc nhận dạng hình ảnh từ camera gắn trên các xe tự hành. Lúc đầu, việclái xe được thực hiện trên phần mềm mô phỏng, thu được các hình ảnh từ 3 camera trái, phải và cameratrung tâm. Chúng tôi cũng thu được dữ liệu 30 lần trong 1 giây từ phanh, vận tốc và góc lái. Dữ liệu thuđược sẽ được thống kê để thu được các tham số đặc trưng của phân bố góc lái, vận tốc trung bình… sauđó cho qua mạng nơ ron CNN để đào tạo và thu được mô hình lái xe tự động. Sau khi đã thu được môhình, chúng tôi lưu vào tệp *.h5 và chạy xe ở chế độ tự lái. Các camera trái phải và giữa cung cấp ảnhtheo thời gian thực. Mô hình sẽ nhận các ảnh này và điều khiển góc lái, vận tốc và sử dụng phanh để tựhành. Chúng tôi đã cho xe chạy thử trên quãng đường dài và thấy xe chạy khá tốt.

1. Đọc trực tuyến

1. Đọc trực tuyến

Tạo bộ sưu tập với mã QR