Researchers improve or commercialize on-board lane departure warning systems

Netease automotive industry 2021-08-05 22:57:37 阅读数:490

researchers improve commercialize on-board board

( The original title : The researchers improved the vehicle lane departure warning system Or commercialize )

Early lane departure warning systems (LDWS) Use more GPS Data, not expensive cameras or maps . According to foreign media reports , Researchers in Minnesota have enhanced and improved the system , Reduces the cost of products that warn drivers of dangerous Lane drift and approaching curves .

According to the American Association of state highway and transportation officials (AASHTO) The data of , Near the main road 60% All fatal accidents are caused by inadvertently deviating from the road or entering another lane . According to the 2018 A study in said , Lane departure accidents are the leading cause of death in traffic accidents in the United States .

Vehicle lane departure warning system (LDWS) Can reduce the severity of all collisions , Including reducing the number of deaths 86%. To achieve zero death (Toward Zero Deaths) The goal is , Minnesota Department of transportation (MnDOT) Committed to developing low-cost and available to all drivers LDWS.

( picture source :MnDOT)

before , from MnDOT And the local road Research Council (LRRB) Joint research projects to create a smart phone based LDWS, Shows great potential in meeting cost and availability standards . But there are still occasional false positives ( The system alarm is activated when there is no lane departure ), Therefore, the system needs to be further investigated and modified .MnDOT and LRRB It is hoped that researchers can solve the false positive anomaly and further develop the on-board system LDWS.

The project aims to improve the on-board equipment developed in the first phase of the research LDWS. The main objective is to address false positives ( No lane departure warning occurred ) Happen , And further develop the system to realize possible commercialization .

To evaluate the vehicle's heading on the road relative to it ( Road reference heading ,RRH) Lateral offset of , Many currently available LDWS The lane position is determined by relying on a camera that detects road markings . however , Markers may degenerate , And in bad weather , Stripes and other markings may also be obscured . Camera based systems may also add GPS technology , Other systems will also use high-resolution digital maps and advanced GPS technology . These systems can be very accurate , But it's also expensive , Usually only for luxury cars .

Researchers intend to use widely available and inexpensive data sources to determine the lateral displacement of vehicles . first , The researchers developed a system algorithm , Will use open source digital maps and standards GPS Information .

In the road test , The first system works well . During the test , The system is used in all parts of the road ( Including curves ) On 200 In the second lane change success An audible alarm was sounded . However , The system is in trial operation , The alarm still sounds when there is no lane departure , The probability of false positives is up to 10%. After inspection , The possible reason is that open source low resolution digital maps contain inherent lateral errors , And it is more obvious in the curved part of the road . The slightly damaged map will turn to LDWS A notice , The vehicle deviated from the curved road , But it's not .

University of Minnesota (University of Minnesota) Professor, Department of electrical engineering, Duluth campus Imran Hayee Express :“ Without using cameras or high-resolution digital maps , The novel algorithm of the system can accurately detect the driver's unintentional lane departure and give an alarm in time . This result paves the way for its integration into smartphone applications , For example, Google maps or Waze.”

To avoid using low resolution RRH Map to evaluate the lateral offset of the vehicle , The researchers developed an enhanced LDWS, The location of a given road can be extracted from the location data of vehicles driving on the same road in the past RRH. The researchers designed a new algorithm , It successfully generates accurate data from the vehicle's past trajectory RRH, Reduce the incidence of false positives to zero . As long as the vehicle has driven on the road , The system can give an accurate alarm , But if it's the first time , The alarm cannot be issued . To solve this problem , Researchers have developed an alternative method , Pass the vehicle to vehicle with other vehicles on the same road (V2V) Communication to determine RRH.

After many field tests on the system using the new algorithm , The system can detect 100 Multiple lane change , And no false positives . assessment V2V The road test of handshake protocol shows that , Receive... From nearby vehicles RRH New data in the vehicle LDWS It can accurately detect lane departure . The project uses standard GPS The technology has successfully developed the vehicle LDWS. Besides , The researchers also designed a system with direct short-range communication V2V Communication protocol , In order to obtain... From another vehicle when needed RRH.

At present, the project has applied for a patent , And the third stage, the implementation stage, has been approved . At this stage , Researchers will develop smartphone applications , As a marketable end product . In the future, the application may also be on Google Maps 、Waza And so on , Allows the user to use a large amount of existing road data .

版权声明:本文为[Netease automotive industry]所创,转载请带上原文链接,感谢。