Several major issues regarding high-precision positioning for autonomous driving

Positioning is the foundation of high-level autonomous driving, but in scenarios such as high-speed NOA and urban NOA, how to stably achieve high-precision positioning under various working conditions will be a problem. A common question is: How high-precision positioning is required for high-speed NOA and urban NOA functions? How high-precision IMU, integrated navigation and how many types of sensors are needed?

With these confusions in mind, the author interviewed many industry experts.

The answer to the first question is relatively consistent: high-speed NOA only needs to achieve lane-level/decimeter-level positioning, and it is enough to be able to identify which lane the vehicle is in; while in urban NOA, because the lanes are relatively narrow, especially at crossroads, At the intersection, lane changes often occur, such as two lanes turning into three or four lanes. At this time, in order to maintain lane keeping without hitting the lane line, positioning accuracy needs to reach centimeter level.

Regarding the second question "How high-precision IMU, integrated navigation and sensors are needed?", the author and many experts in the industry came to the conclusion that it is impossible to determine because the final positioning accuracy is determined by integrated navigation, wheel speed sensors, and high-precision sensors. As long as the accuracy of the final fusion positioning can meet the needs of the precise map and vision, lidar, millimeter wave and other sensors, there are no rigid requirements for the accuracy of each part.

The figure below introduces several components of fusion positioning.

Integrated navigation broadly refers to "the combination of two or more dissimilar navigation and positioning systems", such as data fusion of GNSS, IMU, wheel speedometer, Lidar point cloud and other information through algorithms.

Since GNSS and IMU have a very good complementary effect, that is, GNSS supplements the cumulative error problem of the IMU inertial system, and IMU well compensates for the instability and susceptibility to interference of the GNSS satellite system, this "golden partner" is also called It is called the best integrated navigation solution, so generally when people refer to "integrated navigation", they refer to inertial integrated navigation (GNSS+IMU). Unless otherwise specified, the "combined navigation" mentioned below refers to GNSS+IMU.

This article mainly focuses on the following issues -

one. Why does GNSS have to be dual-frequency?

Currently, all domestic mainstream GNSS/RTK solutions support dual-frequency multi-constellation.

The so-called multi-constellation means that the GNSS receiver can receive signals from different navigation systems such as China's Beidou, the United States' GPS, and Europe's Galileo. Domestic GNSS receivers generally support both Beidou and GPS.

Dual frequency means that each navigation system has two carrier frequency bands. For example, Beidou has two frequency bands, and GPS also has two frequency bands. Dual frequency will lead to a significant increase in hardware and computing power costs.

So, what does dual frequency do?

After communicating with industry experts, the author got the answer: on the one hand, the two carrier frequency bands can be redundant to each other; on the other hand, dual frequency can achieve higher positioning accuracy.

Among the measurement errors of GNSS, the errors caused by the ionosphere account for a large part. The dual-frequency carrier can use the correlation of the ionosphere to the delay of electromagnetic waves of different frequencies to eliminate most of the errors caused by the ionosphere, thus greatly improving satellite positioning. Accuracy, which is beyond the capabilities of single-frequency GNSS.

The founder of an integrated navigation company once publicly mentioned on a forum that the single-frequency RTK positioning accuracy has a 95% probability of falling within a circle with a radius of 0.4 meters (that is, the concept of the accuracy unit CEP). The dual-frequency RTK positioning accuracy can Achieve 95% probability of falling within a circle with a radius of 0.2 meters, which shows the importance of dual frequency.

two. Satellite-based augmentation and ground-based augmentation

In addition to the ionospheric errors mentioned above, GNSS positioning errors include the following: satellite-related satellite orbit errors (ephemeris errors), clock errors of satellite atomic clocks, tropospheric errors, and multipath after carrier reflection Errors caused by effects, receiver clock errors related to the receiver, etc.

△Sources of satellite positioning errors

In order to eliminate these errors as much as possible and improve positioning accuracy, in addition to the dual-frequency carrier band mentioned above, there are other enhancement methods, which can be divided into satellite-based enhancement and ground-based enhancement according to their principles.

Among them, the most widely used and representative enhancement methods are ground-based RTK (Real-Time Kinematic, real-time dynamic positioning), satellite-based PPP (Precise Point Positioning, precise point positioning) and the combination of the two PPP-RTK. The specific technical introduction and advantages and disadvantages are as follows:

△RTK/PPP/PPP-RTK comparison

Information source: Jiuzhang Zhijia compiled based on public information and expert interviews

The figure below compares these three modes from the three dimensions of convergence speed, positioning accuracy, and coverage.

△Comparison of positioning characteristics of RTK, PPP and PPP-RTK

(Source: https://www.sohu.com/a/447492212_120381558)

As a powerful combination of RTK and PPP, PPP-RTK has the advantages of RTK’s high accuracy and fast convergence, as well as the advantages of PPP’s global coverage.

In addition, compared with RTK, PPP-RTK is more in line with functional safety requirements. On the one hand, PPP-RTK is not constrained by the coverage of ground base stations and mobile networks (it can be broadcast through satellites); on the other hand, compared to RTK that packages positioning errors as a "black box" through differential processing, PPP-RTK can An error is modeled and estimated on a global scale. The signal integrity of PPP-RTK allows it to confirm the error status of each signal and identify whether the signal can be converged, usable, and detectable, so as to be more consistent with the function. Requirements for dismantling risk factors item by item.

So, compared with RTK, does PPP-RTK have any cost advantage?

Although PPP-RTK requires a much smaller number of ground base stations (hundreds of domestic base stations can cover it), if you want to use satellites to broadcast signals, you must either rent satellites or launch low-orbit satellites like Spacetime Daoyu. It is understood that the cost of leasing satellites is high (about 10 million to 20 million per satellite per year), and to cover the world, multiple satellites must be rented, and these costs will be included in the service fee.

According to the author’s communication with many industry experts, the cost of PPP-RTK and N-RTK is not much different. If satellite broadcasting is not used, the cost of PPP-RTK will be even lower.

Many experts in the industry believe that PPP-RTK is the future development trend. It is understood that there are already many pre-installation mass production projects under development in China that use PPP-RTK technology.

However, what is interesting is that Qianxun Positioning, which has invested heavily in foundations, is also promoting PPP-RTK to car company customers (according to Qianxun’s official website, Qianxun has built 2,800+ foundation enhancement stations). Is this affected by competition? The opponent's strategy affects you and you have to follow up?

So, after switching to PPP-RTK, will the resources that pioneers invested in ground-based enhancement base stations in the early stage be "wasted"?

The author learned from the communication with a senior practitioner of a head position service provider that although PPP-RTK theoretically does not require so many ground base stations, it will still be affected by the large amount of ground data in terms of convergence speed and position accuracy. Empowerment, PPP-RTK launched based on the existing high-density ground services, its convergence speed is also much faster than similar products of location service providers with less ground density.

three. Can low-orbit satellites improve positioning accuracy?

Some time ago, Geely's Spacetime Daoyu launched nine low-orbit satellites, which are said to be mainly used for high-precision positioning and other functions for intelligent driving. At the same time, Tesla’s “Starlink” also uses low-orbit satellites to cover the world. So, can low-orbit satellites really improve positioning accuracy?

On this issue, feedback from experts is relatively consistent, that is, low-orbit satellites mainly play a communication role and have no substantial role in improving positioning accuracy.

However, although low-orbit satellites cannot improve positioning accuracy, they can indeed enhance coverage of satellite signals, especially in semi-obstructed areas.

A senior practitioner believes that in addition to broadcasting enhanced signals, low-orbit satellites can also use global satellite navigation systems at the same time to broadcast carrier information, thereby increasing the number of satellites searched by GNSS receivers in a certain area. In some semi-obstructed areas, such as under elevated buildings or near windows in buildings, positioning could not be achieved or the positioning accuracy was inaccurate because not enough navigation satellites could be found. With the addition of low-orbit satellites, enough satellites can be found. , thereby enhancing the availability and reliability of satellite navigation.

Four. Key indicators and accuracy requirements of IMU

As the core component of integrated navigation, IMU can provide higher-frequency high-precision signals (generally up to 200Hz). When the GNSS signal is invalid or within the update interval (GNSS signal frequency is 10Hz), it can be used for high-precision dead reckoning. .