CC BY
17
Труды Международного симпозиума «Надежность и качество», 2016, том 2
кл2)
kt2^2)
A./cos(s1.*h)
^ cos(s1.*(h-
=
б2 = зяг1(кл2 к = 1; VI =
z)).*sin(w.*t-k.*x);
v2 = Д.*ехр(з2.*^).*з^^.*^-к.*х); Аналогично рассмотренному ранее способу построим график распространения волны Лява (Рисунок 3) и выгрузим коэффициенты смещения (Рисунок 4).
Необходимо отметить, что форма волны определяется материалом, в котором она распространяется, однако, разработанные модели позволяют подбирать входные параметры моделей, позволяя наиболее приближенно моделировать волны ПАВ, а полученные коэффициенты - наиболее точно вычислять дифференциальные уравнения и с их помощью проводить имитационные эксперименты в Ма^аЬ / З^и^пк.
ЛИТЕРАТУРА
1. Красовский Г.И., Филаретов Г.Ф. Планирование эксперимента.-Минск:ИздательствоБГУ,1982. -302с.
2. Михеев М.Ю., Мещерякова Е.Н. Разработка аналитических моделей сигналов датчиков на ПАВ // XXI век: итоги прошлого и проблемы настоящего плюс. - 2015. - № 4 (26). - С. 222-227.
3. Мурашкина Е.Н., Михеев М.Ю. Имитационное моделирование нейросетевой идентификации сигналов сложной формы // Труды международного симпозиума Надежность и качество. - 2014. -Т. 1. - С. 203206.
4. Мурашкина Е.Н., Михеев М.Ю. Разработка имитационных моделей функционирования подсистемы идентификации и структурирования информации сигналов с датчиков на поверхностно-акустических волнах // Труды международного симпозиума Надежность и качество. - 2015. - Т. 1. - С. 187-190.
5. Мещерякова Е.Н., Михеев М.Ю. Моделирование алгоритма идентификации сигналов датчиков на ПАВ // Теория и практика имитационного моделирования и создания тренажёров. - Пенза. - 2015. - С. 2933.
6. Murashkina E.N. Development sequence diagram of neural network identification of a complex signal using the unified modeling language UML 2.0 // Инновационные информационные технологии. -2014. - № 2. - С. 390-392.
УДК 621.396 Behzadfar Sh.
Samara National Research University, Samara, Russia
INCREASING THE RELIABILITY OF AIRPLANES' ATTITUDE DETERMINATION SYSTEMS WITH GLOBAL NAVIGATION SATELLITE SYSTEMS USING SOFTWARE RECEIVERS
Each year, we hear about tragic plane crash events while landing or take off. Some of these incidents would not happen if the attitude monitoring system on the plane could work accurate at that moment. The most commonly used type of navigation monitoring sensor in each flying object is Inertial Navigation Sensors. These good and anti-jamming devices sometimes suffer from integration drift like small errors in the measurement of acceleration and angular velocity or Gimbal lock. These errors warn engineers to use extra parallel attitude monitoring systems in case they happen. Using Global Navigation Satellite Systems can improve the reliability and quality of attitude estimations. The aim of this research is determining attitude angles by using only GNSS and improving the calculation using LAMBDA method in ambiguity resolution andKalman filter in case of satellite number variation and cycle slips. The overall objective oof this report is to investigate a very accurate solution for GNSS attitude determination system. At the moment, the research result contains only the navigation equations by GNSS and methods to eliminate the effects of different delays or errors in the computation following by LAMBDA method calculations to estimate the ambiguities, and flowchart diagram oof the final simulation code even though some codes have been written according to them but not tested yet. Getting good results and increasing the accuracy of attitude determination by using only GNSS can affect the quality of these computations in navigation monitoring and control systems of airplanes and any other kind of flying object in which only gyroscopes or INS systems have been used.
Key words:
Attitude determination, GNSS, LAMBDA method, Navigation reliability improvement
Introduction
Efficiency of a design is what an engineer should consider all the time. Efficiency can be divided into correct result, reliability, and less cost. To measure the attitude angles in navigation, usually Inertial Navigation Systems (INS) are being used which have correct and precise answers but expensive. Aforementioned, there are some reasons to not trust these methods of navigation all the time and check their results with another method. Recently, defining the attitude of a kinematic system by GNSS is another interesting field of research to solve the navigation system designer's problem. Scientists have tried different methods like using low cost hardware GPS receivers to determine the attitude. This report describes an investigation of a very accurate solution for GNSS attitude determination system. The navigation message in GNSS signal contains all the necessary information to allow users to perform the positioning service. This includes the ephemeris parameters, needed to compute the satellite coordinates with sufficient accuracy, the time parameters and clock corrections, needed to compute satellite clock offsets and time conversions, the service parameters with satellite health information, the Ionospheric parameters model, needed for single-frequency receivers, and the almanacs, allowing computation of the position of all satellites in the constellation , with a reduced accuracy , which is needed for acquisition of the signal by the receiver [1]. Travelling the
signal from space segment to user segment includes some errors which can affect the attitude determination and must be considered such as Satellite clock offset, Ephemeris errors, Ionospheric effects, tropospheric delay, multipath, receiver noise, and some other important errors [2].
Navigation Equations
Typically, code range measurement equation can be written as:
P = p + C(dtr - dts) + T + 1 + £p (1)
p is the geometric range between the satellite and receiver antenna phase centers at emission and reception time. The dtr and dts are the receiver and satellite clock offsets from the GNSS time scale, including the relativistic satellite clock correction. T is the tropospheric delay. 1 is the Ionospheric delay, and £p is error due to different factors like the receiver noise correspond to code range or instrumental delays and multipath effect. The other equation used in the calculation of navigation problems is Carrier phase:
$ = p + c(dtr - dts) + T - 1 + AN + ef (2)
A is the carrier wavelength (m), N is the unknown integer ambiguity (cycles), and €f is error due to different factors like the receiver noise correspond to carrier phase or instrumental delays and multipath effect.
The carrier phase measurements are much more precise than the code pseudorange measurements;
however a receiver can measure only the fractional part of the phase and its variation over time. There is a constant unknown, called initial cycle ambiguity (simply called ambiguity), in every phase measurement (N). Carrier phase measurements become known pseudoranges in case of fixing the integer ambiguities and the determination will be accurate at the level of a few millimeters.
Integer Ambiguity Resolution
In order to determine the baseline vector between two antennas, using the double differences cancels the receiver and satellite clock biases and Ionospheric propagation delay. Comparing the distance between both antennas with the receiver-satellite distance results in cancellation of the tropospheric propagation delay (assuming the elevation angles are the same), m and r are the master and slave receivers' antennas and by adding a new satellite u in the measurement process and double differencing results into:
= wz - Acj>Z = A2PZ + AA2NZ + A2€Z (4)
As a result of double differencing the clock offsets, are cancelled. Doing the same process for code measurements, results in:
A2psu = A2nsu + A2£su (5)
" rmr " Hmr ~ " cpmr (5)
In order to estimate the ambiguities, defining the relationship between the baseline vectors and the double difference equations is necessary. The distance between two antenna receiver in baseline is much smaller than the distance between satellite and each antenna (around 20,200 km), therefore assuming that the signal transmission path to each antenna from a specific satellite are parallel to each other is reasonable. The single difference of geometric range can be written as the inner product of the baseline vector b and the line of sight vector to satellite s, and the double differencing leads to [3]:
A2p™ = b. (em - er) = b.emr (6)
To simplify, the system equation can be written as:
y2(n-l)X1 = B2(n-1)X3b3X1 + ^2(n-1)X(n-1)a(n-1)X1 + e2(n-1)X1
y is the measured double differences' vector, B is the system matrix for the baseline coordinates, containing the differenced Line of sight vectors, b the baseline coordinates' vector, A is the system matrix for the ambiguity set, containing the carrier wavelength, a is the ambiguity set, e is the measurement noise vector. In this resolution process two things have to be done, first Integer ambiguity resolution and then ambiguity resolution quality control. Due to its higher accuracy and better characteristics, LAMBDA method is the chosen method for ambiguity resolution in this research. After implementation of LAMBDA method on the system equation (7), Ambiguity Filter selects the best ambiguity output set from the LAMBDA method and the corresponding baseline is computed in order to be designated merit to each set.
Attitude determination
The coordinates of the baselines in the body frame are constant and known from the beginning. At each epoch the coordinates of baselines in space frame are calculated by LAMBDA method. The next step is to determine the attitude angles by rotation matrix calculation which relate the coordinates in both frames to each other. The rotation matrix is:
Rot(p,9,p) = Rotz(^)Roty(9)Rotx(cp) (11) The transformation from body fixed frame (xyz) to reference space frame NED:
BNED = Rot(ip,Q,y)Bx:
(12)
Where Bxyz and BNED are the matrices containing baseline vectors related to local body fixed frame and reference space frame, respectively. At the end of each epoch by solving the following least square problem [4-5], the rotation matrix can be obtained:
Rot(p, 9, p) = BNEDBXyz(BxyzBXyz)-1 (13)
As a result, attitude angles are equal to:
(14)
(15)
(16)
9 = sin-1(-Rot31) p = sm 11——)
r KcosBJ
ip = sm 1 (——1
r KcosBJ
Solution Implementation
In order to implement all these basic equations in this research the following flowchart was prepared (see figure 1).
Start—►
Data Acquisition
Ref. Receiver Position Init.
Double Differences Formation
Y, 9, 9
Next Repetition m=m+1
Attitude determination
Ambiguity Filter
Figure 1
Basic flowchart for solution
In the first step this flowchart will be tested in a MATLAB code which has been written by the help of the Easy Suite [6]. The next step in this research will be checking the codes by getting data from JAVAD GNSS antennas and its software receiver in Samara National Research University. Although, the whole idea will not stop at this point and there are also other aspects of attitude determination which are likely to be tested and solved, such as:
How to calculate the angles when the number of satellites changes during the observation?
How to estimate the angles in case of cycle slips?
What if the reference satellite changes?
How to check the optimality of the ambiguity filter output?
How can Kalman filter be implemented to improve the accuracy of the float solution?
Conclusion
Aforementioned, for the sake of the mission safety of any flying object, in navigation monitoring, having extra and accurate measuring source is needed due to different errors which can occur during using only Inertial Navigation Sensors. Using LAMBDA method to resolve the ambiguities and implementing ambiguity filter can lead to having a reliable attitude determination unit by GNSS. Also decreasing the amount of hardware used in the navigation monitoring system can be economically and practically helpful. In this research, these goals are aimed to be achieved.
REFERENCES
1. Sanz Subirana, J., Juan Zornoza, J.M., and Hernandez-Pajares, M. (2013). GNSS DATA PROCESSING. Volume I- Fundamentals and algorithms. ESA communications
2. Kaplan E.D and Hegarty C.J. (2006). Understanding GPS: principles and applications. 2nd edition. Artech House Publishers
3. Buist, P. J. and Imparato, D. Undifferenced and Single Differenced GNSS noise analysis through a constrained baseline vector. Available from: https://repository.tudelft.nl/assets/uuid:5e9bb0f0-4b39-4628.../308685.pdf
4. Borre, K., Strang, G., (2012). Algorithms for Global Positioning, Wellesley-Cambridge Press
5. Strang, G., Borre, K., (1997), Linear Algebra, Geodesy, and GPS, Wellesley-Cambridge Press
6. Borre, K., (2009), GPS Easy Suite II: Easy 12: LAMBDA Method, InsideGNSS. p. 48-52. (March/ April). Available from: www.insidegnss.com [Accessed: 20th, 23rd of March 2016]
7. Mikroelektronika, (2012), GPS-Click manual, [Online] Available from: http://www.mikroe.com/click/gps [Accessed: 22nd of March 2016]
YfiK 621.396 Qu.eza.da. Gaibor D.P.
Samara National Research University, Samara, Russia
GNSS - BASED VEHICLE TRACKING AND EMERGENCY RESPONSE TIME SYSTEMS
The aim of this project is to design and develop a vehicle tracking and emergency response time system using multi-constellation GNSS (Global Navigation Satellite System) receiver, with high accuracy, reliability and integrity and to send the position information of vehicle using GSM (Global System for Mobile Communications) modem across the mobile network. Moreover, the proposed system is aimed to monitor vehicle movement and in emergency cases decrease the time response using an integrate system which receive the current position and with this position the system search the nearest response emergency center. Also this system will search the mobile units that are close to the emergency point. All this information will be shown in real time on the monitor agent with a message oof alert. We consider 3 methods to transmit the position received by GNSS antenna: first is using the Internet connection that will be used to save the current vehicle position every second in the data center and to get the location of the object by the monitor agent. It will permit to have a history path oof each vehicle, when the monitor agent could not get the current place. This history information will permit to calculate possible position according the last received position and speed. Second is using a SMS message with parameters like velocity, position and time and thirdly, the GMS modem will call to the emergency center, the telephony system (Asterisk VoIP system) will search the telephone number in the Data Base and it will send an alert to the agent monitor.
Usage of GNSS technology provides precise and reliable positioning information at cheaper cost and also, GSM network for sending vehicle track information is also an efficient means for exchange with the GSM modem between data base and the GNSS receiver in real-time.
Key words:
Multi constellation, GSM modem, real-time, accuracy, time response
Introduction
This project is divided into tree main blocks: the first part is related to GNSS multi constellation receiver with high accuracy, reliability and integrity to compute the current position of the vehicle, second - to the data transmission in real time across the mobile network using a GSM modem and finally to develop a software of 3 layers multi platform with different operation levels such as user application (smart phones, tablets and PC), agent monitor application/software (control of emergency messages and calls) and applications for the mobile units (police, ambulance, fire-fighters). These three points are closely related and work synchronously also they use a free VoIP (Voice Over IP) Service called asterisk.
Most of the available vehicles tracking systems use single constellation GNSS receivers for position computation that is transmitted afterwards over the Internet or mobile network. But the use of single constellation GNSS receiver has limitation in terms of reliability in case of satellite failures. In order to address this problem, I propose in this project to design a Vehicle Tracking and Emergency Response Time system based on multi constellation GNSS receiver that will be integrated with GSM Modem to send positional data over the mobile network. This system configuration improves the reliability as well as integrity of position computation and decreases the time response of emergency center.
Broadly speaking, we can classify vehicle tracking systems in to two types: passive and active tracking system: The passive vehicle tracking system stores the position information of vehicles into the internal memory and it will be downloaded to the data center in the offline for further monitoring. Whereas in active vehicle tracking system, position information will be sent in real time to the data center for online monitoring of vehicles.[1] Some of the active tracking systems use techniques for automated data transmission with ISO standard like SAE (Society of automotive Engineers), J1939 -
ISO (International Organization for Standardization) [5]. This standard provides a high-speed in the transmission of data and control functions in real-time. [5]. Some of these systems can be updated regularly online and setup new trajectory, new features and others parameters. This online dispositive sends all the data in real time and it permits to locate or to follow the trajectory with specific software.
Multi constellation receivers compute position based on the signals from multiple satellite constellations i.e. GPS, Galileo, GLONASS, and others. Multi constellation service provides reliable coverage for position computation anywhere around the world, because failure of few satellites will be compensated by the availability of other satellite constellations. Whereas, in case of single constellation service, failure of single satellite may lead to degradation in the position accuracy. Use of multi constellation has the following advantages: The accuracy of position increases with more visible satellites, the system is more robust, the coverage area is extensive and reliability increases. We can compute more precisely the position with availability of more number of satellites [1].
Architecture
The vehicle tracking and emergency response time system has the next architecture. Firstly, the GNSS receiver in the vehicle gets the data from the GNSS satellites and computes the position. This position information contains latitude, longitude, number of the satellite and time; we will use it in the format established by the National Marine Engineers Association (NMEA), also we will use algorithms to measure the position with the less grade of error, this information will be sent using GSM Modem across the mobile network to the data center where all the information will be saved. Finally, each user could see the tracks of vehicles in devices such as phones or computers, which can be further mapped to electronic maps to ease the user interface. Furthermore, the vehicle-tracking device permits to call and to send information about position to the Emergency Center and the
