Design of GPS Dr Model Design System

Model Design reason:

GPS provides global, around-the-clock, and real-time navigation Advantages. However, when satellite signals such as high-rise buildings, Boulevard, overpasses, and culvert tunnels are "blind spots, GPS cannot provide users with precise navigation information, especially continuous navigation information. Dr (dead-reckoning) uses a direction sensor (Gyroscope) and a speed sensor (mileage meter) to estimate the instantaneous location of a vehicle and realize continuous autonomous positioning. However, as the calculation process is an accumulation process, the error of the direction sensor will accumulate over time. In addition, Dr calculation can only determine the relative position and course of the object. Considering the advantages and disadvantages of GPS navigation and Dr autonomous navigation, the combination of GPS and Dr can complement each other. Dr can effectively solve the "blind zone" of GPS. GPS can suppress the accumulation of Dr errors and solve the Relative Heading problem of dr.
How GPS works:
The GPS system consists of three parts: space, ground support, and user receiver.
We are mainly concerned with the user receiver. GPS receivers can be divided into many types by function purpose, and we are mainly concerned with the navigation type. There are also many ways to use GPS for measurement and navigation.
1. Single Point Positioning: As long as a GPS receiver, it mainly receives the c/a code from the L1 carrier sent by the satellite. The error is generally 0-15 meters, of which about 5 meters has a high probability. Of course, different models may have different precision. However, the maximum error cannot be less than 15 meters.
2. Static Measurement: mainly three GPS devices work at the same time, with a high accuracy. It is mainly used for measurement of control points.
3. Post-processing dynamics (PPK): It is a static measurement. It takes two days for the GPS receiver to work at the same time. The precision is mainly 0-meters, which can basically meet our work needs.
4. Pseudo-range difference (PDK): It is mainly used in coastal areas with a precision of about 1 meter. It uses the difference technology.
5. Carrier Phase Difference (SCN): difference technology is also used. The two GPS receivers need real-time communication, so they have high requirements on receivers. Therefore, the cost is high.
Considering these measurement methods, we may mainly use single point positioning and post-processing dynamics (PPK). Therefore, we mainly understand the working principles of these two methods.
Single point positioning:
GPS receivers receive C/A codes from L1 carriers (only a few satellites exist at l2c and L5 currently, at the same time, it is also affected by SA (US interference added to reduce satellite precision)
Therefore, the precision is low, generally 0-15 meters, and the error is mainly about 5 meters.

Single point positioning chart

The satellite uses L1 carrier to transmit the calendar file to the GPS receiver. The calendar file mainly includes the precise position of each satellite in the WGS-84 coordinate system, the satellite time, and the carrier time of the satellite. The coordinates of Satellite 1 are (x1, Y1, Z1) Satellite 2 (X2, Y2, Z2), satellite 3 (X3, Y3, Z3), and satellite 4 (X4, Y4, z4 ). Based on the formula for finding the distance between two points in space, we only need three satellites to find the position of the GPS receiver in the 84 Coordinate System, however, because the time of the satellite and the time of the GPS receiver may not be synchronized, there may be a clock difference between them, so the other satellite is mainly to find the clock error. That is why single-point positioning requires four satellite signals. This part of the program may be involved mainly in the operations on the data of the calendar file and the clock difference (due to the time not synchronized between the satellite time and the receiver time, read operations on the coordinate data calculated by the GPS receiver. It should not be the focus and difficulty of system design. Moreover, manufacturers selling GPS receivers should also have finished software.
Post-processing dynamics (PPK ):
Two receivers need to work at the same time. One receiver is placed on one known point as the benchmark station and the other is used for measurement. Real-time communication is not required between the two receivers, the working principle of each receiver is equivalent to single point positioning. The distance between the two receivers cannot be too large, generally about 15 kilometers. In addition: the two receivers need to receive signals from five identical satellites at the same time (one of them is mainly used for reference ). Main principle: the receiver of the benchmark station compares the signal received by the receiver and the signal received by the receiver for measurement, and removes the error caused by factors such as electron interference, orbital error, and atmospheric refraction.

Post-processing dynamic diagram

The benchmark station must first start and record data by time period (for example, once a second). In the first second, it detects that the coordinates of the benchmark station are (x1, Y1, Z1), followed by: (X2, y2, Z2), (X3, Y3, Z3), (X4, Y4, Z4 )........ The data measured by the receiver at the second of A is (XA, ya, za ). Because the distance between the two receivers and the distance between the satellite and the receiver, the direct distance between the two receivers can be ignored (15/20200 ), therefore, the error such as the disturbance of the electron, the orbital error and the refraction of the atmosphere can be regarded as the same. Therefore, the error constant is obtained by reference station, which is: Tx = X2-X, Ty = Y2-Y, TZ = Z2-Z. Therefore, the exact coordinate of Point A should add the error constant of point A to the measured value: (xa + Tx, ya + ty, Za + Tz), and B and C. The data of the two receivers must be compared to obtain the exact coordinate value. Therefore, this method is less real-time. This part of the program design is the same as the single-point positioning program design. It mainly compares the data of the two receivers in the calendar file and the clock difference, and operates on a large amount of data. The program should not be the focus and difficulty of system design.
Gyroscope and mileage meter:
Gyroscope can be divided into many types based on working principles and functions, including: fiber optic gyroscope, micro mechanical gyroscope, and electronic gyroscope. The main component used for navigation is the electronic gyroscope. The electronic gyroscope is mainly used to replace the previous magnetic compass. The magnetic compass relies on the physical magnetic medium to obtain the north direction of the magnetic field through the earth's magnetic field. Because it is a physical magnetic medium, it is vulnerable to external influences, however, the electronic gyroscope can overcome this shortcoming. Therefore, the positioning principle of the electronic gyroscope is the same as that of the magnetic compass: the magnetic north direction is obtained through the geomagnetic field, and then the direction of the automobile is obtained, and the angle (magnetic angle) between the automobile exercise direction and the magnetic north direction is recorded ), use this angle to navigate the vehicle. Different types of instruments have different parameters. Gyroscope is usually used together with the mileage meter. The mileage meter obtains the real-time distance of the vehicle through the obtained pulse signal and the time used. In their combination, we may focus mainly on parameters such as angle, angular velocity, acceleration, and number of pulses. If there is a direction (angle), a distance, and a coordinate of the first point, it is mathematical easy to obtain the coordinates of another point, whether scattered or continuous.
This part of the program design should also be related to data operations and serial port problems, as well as the preparation of mathematical formulas, should not be the focus and difficulty of the system design.
GPS receivers have unparalleled advantages, but they also have inevitable defects, such as satellite signal blind spots. The same is true for Dr estimation. It has its own advantages and disadvantages. defect: it requires initial time and error accumulation. Therefore, it is a good method to combine the two instruments in vehicle navigation.
System Design Scheme 1:
Combining Single-point positioning and Dr estimation with data fusion algorithms, the data fusion algorithms currently used in vehicle navigation are mainly Kalman filtering algorithms, kalman filter is divided into two types: Distributed Kalman filter and centralized Kalman filter. Currently, the application and vehicle navigation are mainly distributed Kalman filter. The fusion of single point positioning and Dr estimation through Kalman filtering can effectively prevent GPS data drift and the accumulation of errors during Dr calculation. Basic Principles

Combined use of single point positioning and Gyroscope

At the beginning, the GPS receiver receives the GPS information to obtain the coordinates of the start position (XG, YG, zg). Then, the initial Dr value is used to calculate the value, and then the two values are fused by Kalman filter, that is, obtain the relatively accurate coordinates of the start position (x1, Y1, Z1). The GPS value must be relatively accurate at the beginning (because the Dr value needs to be calculated at the beginning ), in this way, the GPS receiver needs to stay in the same place for a period of time at startup, so that the DR can calculate a value (XD, YD, ZD) with the initial value) (For Dr estimation, it only needs the initial value), and then (XG, YG, zg), (XD, YD, ZD) use Data Fusion to obtain the coordinates (x1, Y1, Z1) with relatively accurate start position ). When the car moves to the second point, the GPS receiver can obtain the coordinates of the second point (xg2, yg2, zg2), through the exact start position (x1, Y1, Z1) initialize Dr calculation, and use the direction and time measured by Dr to calculate the coordinates of the second vertex (xd2, yd2, zd2 ), then, combine the Dr value with the value calculated by the GPS Receiver Using Kalman filter, and obtain the relative exact coordinates (X2, Y2, Z2) of the second vertex ), later. When the GPS receiver fails to receive satellite signals, it relies entirely on Dr for estimation. The initial value is the precise value output by Kalman filter at the previous time. When no satellite signal is available, Dr estimation may result in a cumulative error.
The focus and difficulty of this method should be the fusion algorithm of Kalman filter, which should also be the focus and difficulty of the entire system. Because there are relatively few materials about Kalman filtering, and the knowledge of Kalman filtering is deep, it is difficult to fully understand the essence of the fusion algorithm because it has a wide range of mathematical aspects. At present, only the simplest simulation can be performed (simplifying a large number of considerations). So far, there are still many problems related to Kalman filtering that have not been solved, his understanding of Kalman filtering cannot be completely affirmed. Other program modules that may be involved in this system should have the Coordinate Transformation of different coordinate systems, the reading and storing of a large amount of data, the GPS signal transfer protocol, the data transfer procedures between the instruments, Dr Data calculation data Interoperability with GPS receiver measurement data, the other is the software interface. In these modules, although some functions cannot be fully simulated, these should not be difficult for the system.
Advantages and disadvantages of the solution:
Advantages: the real-time model of the system is good, the cost of the entire system (mainly the instrument) is not high, the system is flexible, and the system error is relatively stable (can effectively suppress the GPS drift phenomenon ).
Disadvantage: the error range may be large (0-15 meters) and data must be processed in real time.
The second design scheme:
Considering that our main application is road measurement and the real-time requirement is not high, we can combine post-processing dynamic (PPK) with Dr estimation.
Working Principle

Combined use of the receiver and gyroscope for Measurement

Assume that points A, B, C, D, E, and F are the locations to be measured, and the satellite signals at a, B, c, and F are good signals (four or more satellites ), however, there is no satellite signal at D, E, or satellite signal difference (the number of satellites is less than 4), while the GPS receiver of the base station remains in the working state. The GPS receivers used for measurement can smoothly obtain the coordinates of A, B, C, and F, which are (XA, ya, za), (XB, Yb, ZB), (XC, YC, ZC), (XF, YF, ZF), when the receiver is at the D point, because there is no satellite signal, the system makes a record (for example, taking time as the record), indicating that there is no satellite signal at the moment or the signal is incomplete. In this case, the system calls the gyroscope navigation and obtains the coordinates of the previous position (point c, initializes the gyroscope with the previous coordinate value. In the process of C-D, the gyroscope records the angle A1 value, and then computes the coordinates of the D point through angle A1 and the mileage meter. Then, the D point is used to calculate the E point until the F point of the satellite signal is rich, here, although the coordinates of the two points D and E can be calculated, the coordinates of the two points de are not accurate enough because the coordinate of point C is not accurate enough, we only need to record the angle from point C to point D and the number of pulses received by the mileage meter, as well as from point D to point F. The data stored by the measurement receiver is (XA, ya, za), (XB, Yb, ZB), (XC, YC, ZC), (XF, YF, ZF), gyroscope angle A1, A2, A3, mileage meter pulse number M1, M2, M3; and then (XA, ya, za), (XB, Yb, ZB) (XC, YC, ZC), and (XF, YF, ZF) are compared with the values measured at the same time point of the base station, get rid of the error caused by common causes (due to the intraday interference, orbital error, and atmospheric refraction), that is, use the PPK technology to obtain the relatively accurate coordinates of the abcf point, then, the gyroscope is initialized with the relatively accurate coordinate of the C point, and the angles A1, A2, A3, M1, M2, and M3 recorded by the gyroscope are used, these quantities are used to export the exact value of the De point.
Of course, the process of measurement data can also be replaced by solution 1, that is, the GPS receiver used for measurement and the gyroscope work together with Kalman filter to obtain the coordinate value of each point, similarly, this process also needs to record the tags without satellite signals, and then compare the merged data with the measurement data of the base station to remove some errors, because the deviation removed by Kalman filter is mainly the signal drift and the gyroscope error accumulation, the causes of these errors and post-processing static (PPK) the causes of the processed error are different (due to the intraday interference, orbital error, and atmospheric refraction. Therefore, the data accuracy principle should be higher, but if they are generated for the same reason, this method will not work.
The system design of this scheme is actually divided into two processes: first, the GPS receiver and gyroscope may involve the following problems: Coordinate Conversion (including the choice of coordinate system ), GPS data reception, analysis of GPS data (determining the number of satellites), interchange of GPS and gyroscope working modes, and receiving of gyroscope data. If filtering is used, Kalman filtering is also considered. Second, the comparison between the base station data and the measurement data may involve the following problems: Analysis of the calendar file, operations on large data volumes, and mutual transmission of data on different machines, the algorithm for finding coordinate points without satellite signals (mainly the compilation of mathematical formulas) and coordinate conversion problems.
Disadvantages of system design: poor operability is not flexible enough (a known point needs to be found for each measurement area). The cost of the system is high (two receivers are required ), poor real-time performance (the data on the two receivers must be compared to obtain more accurate results ). The only advantage is the high precision (1-meters ).
The above description of GPS receivers and gyroscope mainly involves some understanding and understanding of them, and a lot of content cannot be completely affirmed by myself, in addition, the first system scheme is currently more common (mainly used in real-time Vehicle Navigation), and the second scheme is completely self-imagined to meet the measurement needs. How high is its feasibility, I did not know more deeply. The two systems only involve the knowledge they may use and the problems they may encounter. There may be many factors that have not been considered.

Author: Zhang Yu
Reprinted friends must indicate the source. Thank you! Half smoke Ajie
Http://blog.csdn.net/gisfarmer/