In 2009, in the passenger line of Wuhan-Guangzhou, Zhengxi, the application of C3 column control system has put forward higher requirements for high-speed railway train positioning technology. The main technical principles of the C3 column control system are clearly suggested that the operation speed of the train reaches 350 km / h, the minimum tracking interval is 3 min, and 300 km / h and the above-mentioned EMU does not mount the train operation monitoring device, at 300 km / H and above, the CTCS-3 column control system car equipment speed tolerance regulations are speeding 2 km / h alarm, speed 5 km / h triggers common brake, speed 15 km / h triggers emergency brake. These technical principles require that the high-speed railway train operation control system must determine the exact position of the train, including the relevant interval, speed, acceleration and orbite equipment and the distribution of the vehicle equipment resource. This information is to determine if braking measures need to be taken to ensure security intervals. At present, sensors such as gyro, accelerometer, mileper, and GPS receivers have been widely used in train speed registration systems.
1 Development of high-speed railway technology
The development of high-speed railway technology is varied, and various countries use different positioning techniques based on their road conditions, terrain and operational demand. The French AS-Tree system uses Doppler radar for speed-bearing positioning; North American Ares, PTC, PTS system uses GPS (Global Positioning System) to position; European ETCS, Japan CARAT system is positioned with query / transponder and speed sensor; Germany LZB The system uses the railroom cable for train positioning; the US AATC system is positioned with wireless ranging.
According to the China Railway Terrain, the complex conditions of the line and the requirements of the high-speed railway to train positioning technology, the multi-sensor combination positioning is proposed, and the method of combining the GPS / DR / MM combination is used to use the multi-sensor combination positioning technology information complementary characteristics. The resulting information is fused using Kalman filtering to obtain more accurate positioning data than a single sensor.
2 train positioning system solution
The program uses the characteristics of DR autonomous positioning, which ensures that the train can be output anywhere, and the positioning information can be output at any time, and the adoption of the GPS provides the initial position data to DR, and the use of MM satisfies the need for system on the positioning accuracy. The fusion algorithm part uses a federal filtering algorithm to solve the shortcomings of other filtering algorithms calculated the burden and fault tolerance performance.
2.1 selection of train positioning mode
2.1.1 Introduction
GPS is a radio navigation system. As a high-tech applied to the navigation positioning system, there is a worldwide, at any time, any of the continuous three-precision three-dimensional position, speed, and time information of the user under arbitrary weather conditions. When using GPS positioning, you only need to install the receiver on the locomotive, but the positioning accuracy of a local train in the surrounding barrier is affected. Also, GPS is very sensitive to satellite failures, and one satellite has failed, and the GPS performance will deteriorate. Therefore, the GPS positioning information cannot be single as the positional parameters of the column control system.
DR is a comparative classic algorithm used in vehicle positioning navigation. It consists of a sensor and measurement distance of the measurement distance. The mileper is used as a sensor for the measurement distance, and the gyroscope is used as a sensor of the measurement. The milemeter output is a pulse signal. Each of the wheels is turned around, the milemeter outputs a fixed number of pulse signals, and the number of vehicles can be calculated by accumulating the number of pulses of the mileper in a certain period of time, and the distance of the vehicle can be calculated can also calculate the speed of the vehicle. . The gyroscope outputs the angular rate information of the heading angle. The angular rate information integral of the gyro output can receive the relative corner of the train. Compared to GPS, DR can be lost in autonomous positioning, and there is no occlusion of train positioning information caused by occlusion. However, the initial position of the DR system cannot be obtained independently, and the track is calculated as a accumulated process. The measurement errors and calculation errors at different times will accumulate, and the DR error is a divergency process over time.
2.1.2 Positioning Scheme
According to the above analysis of GPS and DR positioning features, this scheme takes multi-sensor combined positioning techniques, namely, various positioning techniques to supplement each other. In the railway line interval, when the GPS information is continuously installed, the GPS information is given to the positioning system, the GPS information is used as the main information, the DR information, and the query transponder information as the verification information, and the three joint filters Give the optimal positioning estimate.
When encountering an obstacle area such as "City Canyon", the GPS signal disappears or weakened, and the DR information is used as the main information. A little bit of GPS failure is prior to the initial position of DR. After the initial location, the mileper and gyroscope can be estimated to be estimated to the next time.
After the train enters the station, the positioning accuracy of GPS and DR is not well expressed due to the small portal spacing, and therefore the query transponder is used to obtain the locating information in the station. At this point, the query of the transponder information as the master information, and the GPS information and DR information are used as the check information.
2.2 Data Fusion Method
The core problem of this program is the design of the positioning algorithm based on data fusion. The data fusion method applied in the field of train speed registration is a judgment detection theory, estimates theory, data association, etc., and the most widely used is the Kalman filtering method in the estimation theory. Compared to other estimation calculations, Kalman filter has a significant advantage: the state space method is designed in the time domain, and the dynamic characteristics of any complex multi-dimensional signal can be described in the time domain, avoiding the signal power in the frequency domain. The trouble of decomposition, the design is simple and easy to use, using the recursive algorithm. So Kalman filter can be applied to any estimate of any stable or non-stationary random vector process, and the resulting estimate is optimal in linear estimation. The filtering algorithms currently developed include linear Karman filtering, extended Kalman filter, and federal karman filtering. The scheme uses the federal Kalman filtering for data fusion.
2.2.1 United Kalman filter model of data fusion
In this filtering algorithm, βm = 0, that is, the main filter has no information input, further optimizing the system, and reduces the amount of operation.
2.2.2 System Filtering Algorithm Steps
(1) Handling the train position information output from the local filter L, and gives a covariance matrix P1 of the status estimation X1 and the estimation error;
(2) Partial filter 2 processes the angular information X2 and train operation distance information of the gyroscope and milemeter output, gives a covariance matrix P2 of state estimation and estimation errors;
(3) Local filter 3 processes information such as the path length of the query transponder output, gives a covariance matrix P3 of the status estimation X3 and the estimation error;
(4) X1, X2, X3, and P1, P2, P3 are sent to the main filter, and the state estimation of the main filter is fused together in formula (1) and formula (2) to obtain global optimal estimates and Co-difference matrix
(5) Resets the status estimate of the three subcaps by the optimal estimate of the main filter. which is
2.2.3 Selection of information allocation parameters
Our other product:
