通过边缘云传感器融合和基于CNN的感知器实时控制车辆
Sumukh Chaurasia1, Parambrata Sanyal1, Gagandeep Kaur1
1Symbiosis Institute of Technology, Nagpur Campus, Symbiosis International (Deemed University), Pune, India.
本研究介绍了一种使用深度学习和物联网 (IoT) 传感器进行自适应车辆控制的混合边缘云系统. 该方法确保安全高效的实时驾驶,即使在不利的条件下.
科学领域:
- 智能运输系统 智能运输系统
- 边缘计算 边缘计算
- 深度学习 (Deep Learning) 是一种深度学习.
背景情况:
- 实时车辆控制对于智能运输系统至关重要,它依赖于快速传感器数据处理来进行感知和决策.
- 目前的系统在适应变化的环境条件和确保运行安全方面面临挑战.
研究的目的:
- 开发一种混合边缘云方法,将深度学习与物联网 (IoT) 传感器融合集成,用于自适应性车辆控制.
- 为了增强对象检测,停止时间预测和在各种驾驶场景下制动控制.
主要方法:
- 使用超声波范围数据与卷积神经网络 (CNN) 融合用于感知任务.
- 实现了混合边缘云架构用于处理和控制.
- 在边缘硬件 (Jetson Nano,Raspberry Pi) 上在正常和模拟的不利驾驶条件下训练和评估CNN模型.
主要成果:
- 在R2 = 0.99 (正常) 和R2 = 0.98 (不利条件) 的情况下实现了高性能.
- 记录了0.0085.5的低平均平方误差 (MSE).
- 证明了低推理延迟 (110-230毫秒),适合实时边缘部署.
结论:
- 混合边缘云方法通过物联网传感器融合和基于CNN的感知实现了自适应的实时车辆控制.
- 该系统在变化的驾驶条件下提高了预测准确性和操作安全性.
- 证实了在智能运输应用的低成本边缘设备上部署深度学习的可行性.
更多相关视频
09:21Real Time Monitoring of Intracellular Bile Acid Dynamics Using a Genetically Encoded FRET-based Bile Acid Sensor
Published on: January 4, 2016
10:28Sensitivity Enhancement of Soft Capacitive Pressure Sensors Using a Solvent Evaporation-Based Porosity Control Technique
Published on: March 24, 2023
相关概念视频
Nuclear Fusion
A helium nucleus has a mass that is 0.7% less than that of four hydrogen nuclei; this lost mass is converted into energy during the fusion. This reaction produces about...
Real Time RT-PCR
The real-time quantification of the number of amplified products is...
Time-Domain Interpretation of PD Control
Consider the example of control of motor torque. Initially, a positive...
Time and frequency -Domain Interpretation of PI Control
Acting as a low-pass filter, the PI controller slows the system's response and extends settling times. This requires...
Time and frequency -Domain Interpretation of Phase-lead Control
The design of phase-lead control involves the strategic placement of poles and zeros to balance steady-state error and system...
Time and frequency -Domain Interpretation of Phase-lag Control
Phase-lag controllers do not place a pole at zero, but instead influence the steady-state error by amplifying any...
