基于电感式磁传感器的车位检测算法设计

Design of Vehicle Detection Algorithm Based on Inductive Magnetic Sensor

截至2018年底,我国汽车保有量已经超过了2亿辆,而城市停车位数量增长远远不及汽车保有量,帮助车主及时获取空余车位分布情况,对缓解由寻找停车位引发的交通堵塞和环境污染十分有效。设计并实现了一种车位检测方法,使用PNI公司的精密三轴电感式磁传感器RM3001,配合PNI12927驱动芯片进行磁场测量,同时将停车位状态转换抽象为隐马尔可夫模型,基于隐马尔可夫模型的特点设计车位检测算法,在MSP430低功耗单片机上实现,从采集到的磁场信号中提取可用特征,判断当前车位是否有车停入或车辆驶出。经过长时间实践验证,该算法获得了较高的准确率,实现了车位检测的目标。

By the end of 2018,China ’ s car ownership has exceeded 200 million,whereas the number of urban parking spaces has grown far less than car ownership. The timely acquisition of spare parking spaces by car owners is very effective in relieving traffic congestion and environmental pollution caused by parking spaces. This paper designs and implements a parking space detection node,which uses PNI’ s precision three-axis inductive magnetic sensor RM3001,with PNI12927 driver chip for magnetic field measurement,meanwhile,we abstract the parking space state into a Hidden Markov Model. The parking space detection algorithm is implemented on the MSP430 low-power single-chip microcomputer based on the characteristics of the Hidden Markov Model. The available features are extracted from the acquired magnetic field signals to determine whether the current parking space has a car parked or the vehicle is driven out. After a long period of time ’ s practical tests,the algorithm achieves a high accuracy and achieves the goal of parking space detection.