KR20200116729A - Method and System for Reduction of Time to First Fix of High Integrity RTK-GNSS - Google Patents

Abstract

The invention relates to a system which reduces the time to first fix (TTFF) of a real time rinematic (RTK)-global navigation satellite system (GNSS) which supports high-precision/high-integrity navigation solutions to unmanned aerial vehicles. According to the present invention, provided are a system and a method for reducing and determining the TTFF of the RTK-GNSS for navigation support of unmanned aerial vehicles. The method comprises the steps of: (a) obtaining an input parameter of an artificial neural network model for estimating a code multipath error; (b) inputting the input parameter into the artificial neural network model; (c) outputting the code multipath error in the artificial neural network model; and (d) correcting a code measurement value using the code multipath error.

Description

Method and System for Reduction of Time to First Fix of High Integrity RTK-GNSS {Method and System for Reduction of Time to First Fix of High Integrity RTK-GNSS}

The present invention relates to a system for shortening the initial preparation time (TTFF, Time To First Fix) of RTK (Real Time Kinematic)-GNSS (Global Navigation Satellite System) that supports high-precision/high-integrity navigation solutions to unmanned aerial vehicles. More specifically, it relates to a system that satisfies the integrity probability and TTFF required for UAV operation by applying a multipath error artificial neural network model construction, multipath error prediction using a neural network, and TTFF analysis based on the integrity required performance.

1 is a diagram showing a carrier-based satellite navigation system.

The RTK-GNSS system based on carrier measurements used for accurate navigation of unmanned aerial vehicles consists of satellites, reference stations, and users. Unlike general code-based GNSS, the carrier-based system can estimate the user's location with cm-class accuracy by differentiating the carrier measurements of the user and the reference station. Due to the characteristics of the carrier, it is not known how many carrier cycles are included in the measured value. This is called an ambiguous constant, and the user's location can be estimated by solving the ambiguous constant.

The initial time it takes to solve the ambiguity constant and calculate the user's location is called TTFF (Time To First Fix), and the ambiguity constant once solved does not change unless a problem occurs in the signal. In the case of a commercial RTK system, it takes about 5 minutes of TTFF, but the ambiguity constant solved in 5 minutes contains a lot of uncertainty, so a large position error may occur and an accident may occur. In order to ensure this, the concept of integrity that detects a large position error and sends a warning to the user must be introduced. Current general RTK systems do not guarantee integrity, and in the case of an unmanned aerial vehicle system using inexpensive equipment, it takes a long TTFF of more than an hour to ensure integrity. In this way, TTFF is determined by the performance required for integrity and the accuracy of the ambiguity constant. Since the accuracy of the ambiguous constant is determined by the multipath error of the GNSS measurement value, TTFF can be reduced by reducing the multipath error. Therefore, a technology (or system) that accurately calculates the ambiguity constant by predicting/mitigating the multipath error and determining the TTFF by reflecting the required integrity required performance must be developed.

In carrier-based RTK-GNSS, the user's location can be estimated by solving the ambiguous constant. To solve the ambiguity constant, a double difference technique is used in which the measurements of the user and the reference station for two satellites are different from each other. The user can measure the code measurement value and the carrier measurement value from the GNSS satellite, and each measurement value is expressed as Equation 1 below.

here,

ρ : code measurement [m]

φ : carrier measurement value [cycle]

λ : Carrier wavelength [m/cycle]

r : actual distance between receiver and satellite [m]

b : receiver clock error [m]

B : Satellite clock error [m]

I : delay error due to ionosphere [m]

T : delay error due to convection layer [m]

ν : Multipath error ~ N (0 ,σ ² ) [m]

N : Ambiguous constant [cycle]

Subscript: means user k

Superscript: means satellite i

Equation 2 below represents a single difference measurement in which the measurement values of the user and the reference station are different from each other, and Equation 3 represents a double difference measurement in which the measurement values of the two satellites are differentiated from each other.

와

는 무시할 수 있을 만큼 작아지게 되고, 측정치 간에 독립적인 오차요소인 다중경로 오차가 모호정수 및 상대위치만 남게된다. 따라서 이중차분 코드측정치와 이중차분 반송파 측정치를 이용하여 모호정수 및 상대위치를 결정한다. 반송파의 다중경로 오차는 코드 다중경로 오차에 비해 무시할 수 있을 만큼 작기 때문에 모호정수의 정확도는 코드 다중경로 오차(ν _ρ )에 의해 결정된다고 할 수 있다. 다중경로 오차는 주변 장애물에 의해 반사된 신호가 원래의 신호와 간섭을 일으켜 발생하는 지연 오차를 의미한다. GNSS 위성은 일정 주기를 갖고 지구를 공전하고 있으므로 고정된 수신기에 대해 다중경로 오차는 주기적으로 반복되며 주변 환경에 영향을 많이 받는다. 이러한 특성을 가진 다중경로 오차를 완화시키기 위해 1)수신기에서 하드웨어/소프트웨어 적으로 완화하는 방법과 2)발생한 다중경로 오차를 추정하여 보상해주는 방법이 있다. 수신기의 하드웨어 관점에서 초크링(Choke Ring) 또는 배열 안테나를 사용 하는 방법이 있고 소프트웨어 관점에서 협대역 상관기(Narrow Correlator) 안테나를 사용하는 방법이 있다. 하지만 위의 방법들은 비싸고 무겁기 때문에 무인기에 적용하기에는 무리가 있다. 다중경로 오차를 추정하여 보상하는 방법에는 전날 데이터에 적응필터를 이용하여 추정하는 방법이 있다. 하지만 적응필터 방법의 경우 고정된 수신기에 대해서만 사용 가능하다. 따라서 비행환경에 따라 다중경로 오차가 변화하는 무인기에 사용하기에는 한계가 있다.In the process of double difference, the error is due to the correlation between the measured values.

Wow

Becomes negligibly small, and the multipath error, which is an independent error factor between the measured values, remains only the ambiguous constant and the relative position. Therefore, the ambiguity constant and the relative position are determined using the double difference code measurement value and the double difference carrier measurement value. Since the multipath error of the carrier is negligible compared to the code multipath error, the accuracy of the ambiguous constant can be said to be determined by the code multipath error ( ν _ρ ). Multipath error refers to a delay error that occurs when a signal reflected by a nearby obstacle interferes with the original signal. Since the GNSS satellite orbits the earth with a certain period, the multipath error is periodically repeated for a fixed receiver and is greatly affected by the surrounding environment. In order to mitigate the multipath error with this characteristic, there are 1) a hardware/software mitigation method in the receiver and 2) a method of estimating and compensating the multipath error generated. There is a method of using a choke ring or an array antenna from a hardware perspective of a receiver, and a method of using a narrow correlator antenna from a software point of view. However, the above methods are expensive and heavy, so it is difficult to apply them to UAVs. As a method of estimating and compensating for a multipath error, there is a method of estimating the data of the previous day using an adaptive filter. However, the adaptive filter method can only be used for a fixed receiver. Therefore, there is a limit to use in an unmanned aerial vehicle whose multipath error varies depending on the flight environment.

KR 10-0980762 B1

The present invention was invented to solve the above-mentioned problems, and an object of the present invention is to provide a system and method for shortening and determining the initial preparation time (TTFF) of an RTK-GNSS for navigation support of a UAV.

In order to achieve such an object, the method for shortening the initial preparation time of the highly secure RTK-GNSS according to the present invention includes: (a) an input parameter of an artificial neural network model (hereinafter referred to as'artificial neural network model') for code multipath error estimation. Obtaining a; (b) inputting the input parameter into the artificial neural network model; (c) outputting a code multipath error in the artificial neural network model; And (d) correcting the code measurement value using the code multipath error.

The input parameter may include a parameter indicating a direction in which the receiver looks at the satellite; Signal to noise ratio (SNR); And it may include at least one of the inclination of the antenna installed in the UAV.

Between the steps (c) and (d), (c1) collecting a code multipath error output from the artificial neural network model for a preset time; (c2) obtaining an average of the collected code multipath errors, and estimating the average as a code multipath error.

The code multipath error in step (d) may be the code multipath error estimated in step (c2).

The artificial neural network model may be formed by learning by a layer arranged in the same manner as the artificial neural network model.

The step (a) may include calculating a parameter indicating a direction in which the receiver looks at the satellite from the GPS data received from the satellite, and obtaining the parameter as an input parameter of the artificial neural network model.

The step (a) may include calculating a signal to noise ratio (SNR) from the GPS data received from the satellite by the receiver, and obtaining this as an input parameter of the artificial neural network model.

The step (a) may include calculating a slope of an antenna installed in the UAV from the data acquired by the sensor unit, and obtaining this as an input parameter of the artificial neural network model.

In the step (d), the final code measurement value may be corrected and calculated by subtracting the code multipath error of the step (d) from the previously measured code measurement value.

According to another aspect of the present invention, a system for shortening an initial preparation time of a highly secure RTK-GNSS includes at least one processor; And at least one memory storing a computer-executable instruction, wherein the computer-executable instruction stored in the at least one memory, by the at least one processor, (a) for estimating a code multipath error Acquiring an input parameter of an artificial neural network model (hereinafter referred to as “artificial neural network model”); (b) inputting the input parameter into the artificial neural network model; (c) outputting a code multipath error in the artificial neural network model; And (d) correcting the code measurement value using the code multipath error.

The system for shortening the initial preparation time of the highly secure RTK-GNSS may further include a receiver for receiving GPS data for calculating an input parameter of the artificial neural network model from an artificial satellite.

The input parameter may include a parameter indicating a direction in which the receiver looks at the satellite; And one or more of a signal to noise ratio (SNR).

The system for shortening the initial preparation time of the high safety RTK-GNSS may further include a sensor unit for calculating an inclination of an antenna installed in the UAV.

According to another aspect of the present invention, a computer program for shortening the initial preparation time of the highly secure RTK-GNSS is stored in a non-transitory storage medium, and by a processor, (a) an artificial neural network model for estimating code multipath errors Obtaining an input parameter of (hereinafter referred to as'artificial neural network model'); (b) inputting the input parameter into the artificial neural network model; (c) outputting a code multipath error in the artificial neural network model; And (d) correcting the code measurements using the code multipath error.

According to the present invention, there is an effect of providing a system and method for shortening and determining the initial preparation time (TTFF) of the RTK-GNSS for navigation support of the UAV.

1 is a diagram showing a carrier-based satellite navigation system.
2 is a flow chart of a multipath error mitigation algorithm.
3 is a diagram illustrating a change in antenna gain according to an antenna inclination.
4 is a diagram showing the configuration of a system for shortening the initial preparation time of the high safety RTK-GNSS.
5 is a diagram for estimating a multipath error according to an environment.
6 is a diagram showing a first result of multipath error standard deviation according to elevation angle.
7 is a diagram for estimating a multipath error according to an antenna tilt.
8 is a view showing a second result of the multipath error standard deviation according to the elevation angle.
9 is a diagram showing multipath error standard deviation and integrity probability-based TTFF.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings. Prior to this, terms or words used in the specification and claims should not be construed as being limited to their usual or dictionary meanings, and the inventors appropriately explain the concept of terms in order to explain their own invention in the best way. Based on the principle that it can be defined, it should be interpreted as a meaning and concept consistent with the technical idea of the present invention. Accordingly, the embodiments described in the present specification and the configurations shown in the drawings are only the most preferred embodiment of the present invention, and do not represent all the technical spirit of the present invention, and thus various alternatives that can be substituted for them at the time of application It should be understood that there may be equivalents and variations.

2 is a diagram illustrating a flowchart of a multipath error mitigation algorithm, and FIG. 3 is a diagram illustrating an antenna gain change according to an antenna slope.

The present invention uses an artificial neural network to estimate and compensate for code multipath errors in real time.

Referring to FIG. 2, an artificial neural network model for estimating a code multipath error is constructed by training using previously collected data (S210), and a multipath error of the current measurement is estimated in real time using the learned model. The multipath error is alleviated by performing (S230) correction (S240). Since it is a model for estimating multipath errors, the output of the artificial neural network must be a multipath error, and the input must be a parameter that can well represent the characteristics of the multipath error and can be used in real time. Therefore, as inputs of the model, parameters indicating the direction from which the receiver looks at the satellite, signal to noise ratio (SNR), and the slope of the antenna are used as inputs.

The parameters indicating the direction in which the receiver looks at the satellite may be, for example, an elevation angle and an azimuth angle, and this data can be calculated and used in real time from GPS data received from the satellite in the receiver. Since the reflected signal changes according to the relative position between the receiver and the satellite, not the absolute position of the satellite, the multipath error can give information about the environment around the receiver to the artificial neural network model.

SNR is the ratio of the signal strength and noise strength received by the antenna, and can be calculated and used in real time from GPS data received from a satellite in a receiver. When the reflected signal directly interferes with the signal, not only the multipath error but also the SNR changes. At this time, for the difference in propagation distance between the direct signal and the reflected signal, the multipath error and the change in SNR are in an in-phase relationship. Although it is difficult to derive an exact relationship between multipath error and SNR, it can be seen that SNR has a strong correlation with multipath error. Therefore, although it cannot be distinguished by the elevation and azimuth angles, which are not very sensitive to the movement of the UAV, the environment around the UAV that changes little by little can be distinguished through SNR.

The tilt of the antenna can be used in real time through the sensor unit 20 (refer to FIG. 4) of the UAV. Such a sensor unit 20 (refer to FIG. 4) may include an inertial measurement unit (IMU) sensor. Since the gain of the antenna has different values depending on the direction around the antenna, even the same signal differs in the received signal depending on the direction in which the signal is received. As shown in FIG. 3, since the UAV is tilted when it moves, the antenna attached to the UAV is also tilted and the antenna gain pattern changes. Therefore, a difference occurs in the received signal and the multipath error also changes.

In the step of constructing an artificial neural network model for estimating the code multipath error (S210), parameters indicating the direction from which the receiver looks at the satellite, such as i) elevation and azimuth, ii) SNR, iii) tilt of the antenna, etc. An artificial neural network model is trained by taking a plurality of data sets to be included as inputs and using a code multipath error calculated for each data set as a predetermined output.

Thereafter, data such as parameters indicating the direction in which the receiver looks at the satellite, SNR, and tilt of the antenna are acquired in real time and used as an input of the learned and constructed artificial neural network model (S220).

In the artificial neural network model, code multipath errors are output for such inputs. The code multipath error output as described above is collected for a predetermined time using an artificial neural network model, and the average is averaged to finally estimate the code multipath error. The average is obtained, as an embodiment, of obtaining a root mean square (RMS) value of the collected code multipath error values.

가 된다. 여기서 i는 인공위성의 번호, j는 무인기에 장착된 수신기의 번호이다.In this way, the final estimated code multipath error is

Becomes. Where i is the number of the satellite and j is the number of the receiver mounted on the UAV.

로부터, 이와 같이 최종적으로 추정된 코드 다중경로 오차

를 제거하는 보정을 하여(S240) 이와 같이 보정된 값인

을, 보정된 코드 측정치

로 결정하게 된다(S240).In addition, in Equation 4, the previously measured code measurement value

From, this finally estimated code multipath error

Correction to remove is performed (S240), which is the corrected value

, Corrected code measurements

It is determined as (S240).

가 무시할 수 있을 정도로 작은 값이므로, 단계 S240에서 보정된 값

과 기 측정된 반송파 측정치

을 대입하면, 수학식 3의 최종식에서 모호정수

를 결정할 수 있다.In the final equation of Equation 3 representing the carrier measurement value, the multipath error

Is a negligibly small value, so the corrected value in step S240

And measured carrier wave measurements

Substituting for, the ambiguity constant in the final expression of Equation 3

Can be determined.

에서 모호정수 항인

을 제거함으로써 바로 수신기 j와 위성 i 사이의 실제 거리인

를 구할 수 있게 되고, 이와 같은 방식으로

를 4개의 위성에 대하여 구하게 되면 무인기의 현재 위치를 계산해 낼 수 있다.In addition, since the ambiguous constant determined as described above does not change unless a problem occurs in the signal, the carrier measurement value of Equation 3 is used to determine the position of the UAV during future flight.

In ambiguous integer

By removing Δ, the actual distance between receiver j and satellite i

Can be obtained, and in this way

If you obtain the for 4 satellites, you can calculate the current position of the drone.

As described above, the code multipath error output using the artificial neural network model is collected for a predetermined time, averaged, and finally estimated as the code multipath error. The time required for this is called Time To First (TTFF). Fix). As described above, calculating the average may mean obtaining a root mean square (RMS) value of the collected code multipath error values as an embodiment.

In the conventional method of estimating the code multipath error, the deviation is very severe, so it took a long time to collect the code multipath error data, but a method of estimating the code multipath error using an artificial neural network model as in the present invention. The difference in code multipath error values output from the artificial neural network model is small, and thus reliable code multipath error can be finally estimated by collecting code multipath error data for a much shorter time.

4 is a diagram showing the configuration of a system 100 for shortening the initial preparation time of the highly secure RTK-GNSS.

The system 100 for shortening the initial preparation time of the highly secure RTK-GNSS, the receiver 300 and the sensor unit 20 are mounted on an unmanned aerial vehicle.

The receiver 300 serves to receive GPS data from the satellite 10 and transmit it to the system 100 for shortening the initial preparation time of the highly secure RTK-GNSS.

The system 100 for shortening the initial preparation time of the highly secure RTK-GNSS includes a processor 110, a nonvolatile storage unit 120 for storing programs and data, a volatile memory 130 for storing programs being executed, and a receiver 300. ) And a communication unit 140 for performing communication, and a bus that is an internal communication path between these devices. As a running program, there may be a device driver, an operating system, and various applications. Although not shown, the electronic device includes a power supply unit such as a battery.

Applications include a code multipath error estimation application 220, a GPS data processing application 230, and the like.

The code multipath error estimation application 220 includes an artificial neural network model for code multipath error estimation as described above with reference to FIG. 2, and estimates a code multipath error therefrom.

The GPS data processing application 230 calculates input data into an artificial neural network model for estimating a code multipath error as described above with reference to FIG. 2. That is, from the GPS data received from the receiver 300, a parameter indicating a direction from which the receiver looks at the satellite, such as elevation and azimuth, or a signal to noise ratio (SNR) is calculated, and a code multipath error estimation application ( 220) to the artificial neural network model for multipath error estimation.

In addition, the sensor control driver 210 calculates the slope of the antenna based on the data received from the sensor unit 20 and transmits it to an artificial neural network model for estimating a code multipath error. As described above, this sensor unit may include an inertial measurement unit (IMU) sensor. Alternatively, without the sensor control driver 210, the sensor unit 20 directly calculates the slope of the antenna using the detected signal and transmits it to the code multipath error estimation application 220, so that the code multipath error estimation application 220 It can be used as an input to the artificial neural network model.

5 is a diagram for estimating a multipath error according to an environment, FIG. 6 is a diagram showing a first result of a multipath error standard deviation according to an elevation angle, and FIG. 7 is a diagram for estimating a multipath error according to an antenna slope , FIG. 8 is a diagram showing a second result of the multipath error standard deviation according to the elevation angle.

5 is an example of estimating a code multipath error by an artificial neural network model for code multipath error estimation of the code multipath error estimation application 220 described with reference to FIG. 2.

The first example collected data for two days for two environments (type A and type B) and trained a neural network model using data from one day. It is possible to check whether the model from which various environments have been trained can estimate the multipath error by distinguishing the environment around the UAV well.

The results of testing using data of a different day for each environment are shown in FIG. 5. Blue and black represent actual data and red represent estimated values. As shown in FIG. 5, it can be seen that the multipath error is estimated by distinguishing the two environments well from each other. Statistically looking at the above results, in the case of type A, the standard deviation of the unmitigated multipath error is about 0.19m, and the standard deviation of the multipath error relaxed by compensating for the multipath error using an artificial neural network is about 0.13m, which is 31%. The multipath error of is reduced. Type B is also found to be 0.20m when it is not mitigated, and 0.14m when it is mitigated.

A graph of the first result of the standard deviation of the multipath error according to the elevation angle of the satellite is shown in FIG. 6, and it can be seen that the standard deviation decreases for most satellite elevation angles.

The second example shows whether the multipath error can be well estimated according to the antenna tilt. In the same environment, data for two cases at 0 and 30 were collected for two days, and then a neural network was trained using the data for one day. The results of testing using data from another day are shown in FIG. 7. Blue and black represent actual data, and red represent estimated values.

As shown in FIG. 7, it can be seen that multipath errors are well classified and estimated for each angle. In the case of 0 degrees, the standard deviation of the multipath error was reduced by about 42% from about 0.16m to 0.09m, and in the case of 30 degrees, it could be confirmed that it decreased by about 30% from 0.19m to 0.13m.

9 is a diagram showing multipath error standard deviation and integrity probability-based TTFF.

)과 다중경로 오차의 표준편차에 대한 TTFF를 나타내고 있다. 도 9에서 보이듯이 다중경로 오차의 표준편차가 작을수록 무결성 요구조건이 완화될수록 TTFF가 감소하는 것을 볼 수 있다. 시스템에서는 도 9와 같이 2단계서 완화시킨 다중경로 오차의 표준편차와 운영하고자 하는 시스템의 무결성 요구조건을 기반으로 TTFF를 설정한다. 무결성 요구조건은 무인기 임무에 따라 기존에 미리 설정하는 설계 파라미터이다. 도 9는 RTK-GNSS 시스템이 반송파 이중차분 측정치와 TTFF 시간동안 필터링되는 Geometry-free 측정치를 기반으로 운영되는 것을 가정하였다.TTFF is determined based on the estimated code multipath error information and the integrity information required by the system. 9 shows the integrity probability (

) And the TTFF for the standard deviation of the multipath error. As shown in FIG. 9, it can be seen that as the standard deviation of the multipath error is smaller, the TTFF decreases as the integrity requirement is relaxed. In the system, as shown in FIG. 9, the TTFF is set based on the standard deviation of the multipath error relaxed in step 2 and the integrity requirement of the system to be operated. The integrity requirement is a design parameter previously set according to the UAV mission. FIG. 9 assumes that the RTK-GNSS system is operated based on the carrier dual difference measurement and the geometry-free measurement filtered during the TTFF time.

10: satellite
20: IMU sensor unit
100: High safety RTK-GNSS shortened initial preparation time system
210: sensor control driver
220: code multipath error estimation application
230: GPS data processing application
300: receiver
500: drone

Claims (14)

As a method of shortening the initial preparation time of high safety RTK-GNSS,
(a) obtaining an input parameter of an artificial neural network model (hereinafter referred to as'artificial neural network model') for estimating a code multipath error;
(b) inputting the input parameter into the artificial neural network model;
(c) outputting a code multipath error in the artificial neural network model; And
(d) correcting the code measurement value using the code multipath error
A method of shortening the initial preparation time of the high safety RTK-GNSS including a. The method according to claim 1,
The input parameter is,
A parameter indicating a direction in which the receiver looks at the satellite;
Signal to noise ratio (SNR); And
Tilt of the antenna installed on the drone
Method for shortening the initial preparation time of the high safety RTK-GNSS, characterized in that it comprises at least one of. The method according to claim 1,
Between step (c) and step (d),
(c1) collecting a code multipath error output from the artificial neural network model for a preset time;
(c2) obtaining an average of the collected code multipath errors, and estimating the average as a code multipath error.
Method for reducing the initial preparation time of the high safety RTK-GNSS, characterized in that it further comprises. The method of claim 3,
The code multipath error in step (d) is,
Is the code multipath error estimated in step (c2)
A method of shortening the initial preparation time of the high safety RTK-GNSS, characterized by. The method according to claim 1,
The artificial neural network model,
Learned and formed by layers arranged in the same way as the artificial neural network model
A method of shortening the initial preparation time of the high safety RTK-GNSS, characterized by. The method according to claim 1,
The step (a),
Calculating a parameter indicating a direction in which the receiver looks at the satellite from the GPS data received from the satellite, and obtaining this as an input parameter of the artificial neural network model
Method for shortening the initial preparation time of the high-safety RTK-GNSS comprising a. The method according to claim 1,
The step (a),
Calculating a signal to noise ratio (SNR) from GPS data received from a satellite by a receiver, and obtaining this as an input parameter of the artificial neural network model
Method for shortening the initial preparation time of the high-safety RTK-GNSS comprising a. The method according to claim 1,
The step (a),
Calculating the slope of the antenna installed in the UAV from the data acquired by the sensor unit, and obtaining this as an input parameter of the artificial neural network model
Method for shortening the initial preparation time of the high-safety RTK-GNSS comprising a. The method according to claim 1,
In step (d), the final code measurement value,
It is calculated by subtracting the code multipath error of the step (d) from the previously measured code measurement value.
A method of shortening the initial preparation time of the high safety RTK-GNSS, characterized by. As a system to shorten the initial preparation time of high safety RTK-GNSS,
At least one processor; And
Including at least one memory for storing computer-executable instructions,
The computer-executable instruction stored in the at least one memory, by the at least one processor,
(a) obtaining an input parameter of an artificial neural network model (hereinafter referred to as'artificial neural network model') for estimating a code multipath error;
(b) inputting the input parameter into the artificial neural network model;
(c) outputting a code multipath error in the artificial neural network model; And
(d) correcting the code measurement value using the code multipath error
A system that shortens the initial preparation time of the highly secure RTK-GNSS to ensure that it is implemented. The method of claim 10,
Receiver for receiving GPS data for calculating input parameters of the artificial neural network model from an artificial satellite
High safety RTK-GNSS initial preparation time reduction system, characterized in that it further comprises. The method of claim 11,
The input parameter is,
A parameter indicating a direction in which the receiver looks at the satellite; And
Signal to noise ratio (SNR)
High safety RTK-GNSS, characterized in that it comprises at least one of the initial preparation time reduction system. The method of claim 10,
Sensor unit for calculating the inclination of the antenna installed in the UAV
High safety RTK-GNSS initial preparation time reduction system, characterized in that it further comprises. As a computer program to shorten the initial preparation time of high safety RTK-GNSS,
It is stored in a non-transitory storage medium, and by the processor,
(a) obtaining an input parameter of an artificial neural network model (hereinafter referred to as'artificial neural network model') for estimating a code multipath error;
(b) inputting the input parameter into the artificial neural network model;
(c) outputting a code multipath error in the artificial neural network model; And
(d) correcting the code measurement value using the code multipath error
A computer program to shorten the initial preparation time of a highly secure RTK-GNSS that contains instructions to run.

Images