A method for positioning and tracking trains on a curve

Abstract

The present invention provides a positioning and tracking method for trains on a bend, characterized in that: the bend is covered by a base station signal for train tracking, a head terminal and a tail terminal are respectively arranged at the head and the tail of the train, and both the head terminal and the tail terminal are configured with multi-antenna arrays; the base station simultaneously sends downlink signals to the head terminal and the tail terminal in real time; three measurement values of TOA, AOA and Doppler frequency shift of the head and the tail are respectively calculated; the UKF algorithm or the PF algorithm is combined with the cooperative positioning idea to simultaneously obtain the positioning coordinates of the head and the positioning coordinates of the tail at the nth moment. The positioning and tracking method of the present invention solves the feasibility problem of single-base-station positioning and tracking of trains on a bend, reduces the construction cost, and further improves the positioning accuracy.

Description

Technical Field

[0001] The present invention relates to the technical field of rail transit, and particularly to a method for positioning and tracking a train on a curve. Background Art

[0002] In the research on the current state-of-the-art train positioning technologies, the train positioning and tracking method based on the new generation of railway mobile communication system can obtain the estimation of the states including the train position, speed, and running direction by measuring wireless signal parameters (TDOA, TOA, RSSI, AOA). However, the existing train positioning and tracking method based on wireless communication requires at least two or more base stations to achieve, resulting in a large signaling overhead of the network, a heavy network burden, and a high construction cost.

[0003] The above train positioning method mainly refers to the train head. In the conventional train formation operation mode, the position information of the train head is usually used as the train positioning information, and only the position tracking of the train head is required for train positioning. However, with the rise of the virtual coupling technology based on vehicle-to-vehicle communication, the tracking running interval distance between trains is extremely small. To ensure the train operation safety and control the running interval between trains in real time and accurately, it is necessary to track and obtain the position information of both the train head and the train tail simultaneously. When the train is running on a straight track, with the help of the train length information and the running direction information, it is relatively easy to establish the geometric relationship between the position information of the train head and the train tail, so as to calculate the position information of the train tail through the obtained position information of the train head. However, when the train is running on a curve, the above position information relationship between the train head and the train tail is destroyed, and the position information of the train tail cannot be directly deduced from the position information of the train head. And by increasing the number of base station configurations to respectively position and track the train head and the train tail, the network burden will be further increased and the line construction cost will be raised. Summary of the Invention

[0004] Aiming at the problems in the background art, the present invention provides a method for positioning and tracking a train on a curve, so as to solve the problem of how to simultaneously position and track the train head and the train tail on a curve only through a single base station configuration.

[0005] To achieve the purpose of the present invention, the present invention provides a method for positioning and tracking a train on a curve, and the innovation lies in that: the curve is covered by a base station signal for train tracking, and a positioning signal communication terminal is respectively arranged at the train head and the train tail of the train. The positioning signal communication terminals arranged at the train head and the train tail are respectively denoted as the train head terminal and the train tail terminal, and both the train head terminal and the train tail terminal are configured with multi-antenna arrays;

[0006] The positioning and tracking method includes:

[0007] (1) The base station simultaneously sends downlink signals to the head terminal and the tail terminal in real time; the downlink signals include pilot signals and synchronization signals;

[0008] (2) Calculate the three measurement values of TOA, AOA, and Doppler frequency shift of the head according to the downlink signals obtained by the head terminal; at the same time, calculate the three measurement values of TOA, AOA, and Doppler frequency shift of the tail according to the downlink signals obtained by the tail terminal;

[0009] (3) Use the UKF algorithm or the PF algorithm to obtain the positioning coordinates of the head at time n and the positioning coordinates of the tail

[0010] In step (3), the state vector equations involved in the UKF algorithm or the PF algorithm are: s[n] = f(s[n - 1]) + q[n], and the measurement equation is: G[n] = h(s[n]) + u[n];

[0011] Among them, n is a certain moment in the discretized sampling time, s[n] is the state variable at time n, and its definition is:

[0012]

[0013] Among them, and are the abscissa and ordinate of the head at time n respectively, v[n] is the train speed at time n, and φ 1 [n] is the running direction angle of the head at time n, and are the abscissa and ordinate of the tail at time n respectively;

[0014] f(s n-1 ) is the non-linear state transition function at time n - 1, and the definition of f(s n-1 ) is:

[0015]

[0016] Among them, s[n - 1] is the state variable at time n - 1; Δt is the sampling time interval; the definition of s[n - 1] is:

[0017]

[0018] Among them and are the abscissa and ordinate of the head at time n - 1 respectively, v[n - 1] is the train speed at time n - 1, and φ 1 [n - 1] is the running direction angle of the head at time n - 1, and They are respectively the abscissa and ordinate of the rear of the vehicle at time n-1; φ 2 [n-1] is the running direction angle of the rear of the vehicle at time n-1;

[0019] Among them, the running direction angle φ 1 [n] of the front of the vehicle and the running direction angle φ 2 [n] of the rear of the vehicle are determined by the following formula:

[0020]

[0021] Among them, L is the formation length of the train, and R is the minimum curve radius of the curve;

[0022] q[n] is the system noise vector at time n, and the q[n] follows Gaussian white noise with a mean of zero, and the q[n] is a set value;

[0023] G[n] is the measurement variable at time n, and the definition of G[n] is:

[0024]

[0025] Among them, and are respectively the three measurement values of TOA, AOA and Doppler frequency shift of the front of the vehicle at time n, and are respectively the three measurement values of TOA, AOA and Doppler frequency shift of the rear of the vehicle at time n;

[0026] h(s[n]) is the non-linear measurement function at time n, and the definition of h(s[n]) is:

[0027]

[0028] Among them, B is the position coordinate of the base station, and the B=(B x , B y ); c is the speed of light; λ is the wavelength of the signal; the symbol ||||2 represents the second-order norm;

[0029] u[n] is the measurement noise vector at time n, and the q[n] follows Gaussian white noise with a mean of zero, and the q[n] is a set value.

[0030] As an optimization, the communication among the front-end terminal, the rear-end terminal and the base station adopts a 5G-R network communication system.

[0031] As an optimization, the least squares algorithm is used to update the T 1 [n] in the step (iii) to obtain the corrected front-end positioning coordinates at time n The least squares algorithm is used to update the T 2[n] is updated to obtain the corrected tail positioning coordinates at time n

[0032] The principle of the present invention is as follows:

[0033] Since a train running on a curve cannot, like a train running on a straight track, calculate the positioning information of the tail or head only through geometric relationships based on the positioning information of the head or tail, in order to obtain the positioning information of both the head and the tail simultaneously, the present invention is equipped with positioning signal communication terminals at both the head and the tail of the train on a curve respectively, so as to obtain the parameter data related to positioning of the head and the tail simultaneously and in real time, and thus obtain the positioning information of the head and the tail in real time according to these parameter data. However, whether it is to track the positioning of the head or the tail alone, at least more than 2 base stations are required to achieve it. Moreover, to simultaneously position the head and the tail of a train running on a curve, in the prior art, the configuration of a single base station is very difficult to achieve. In the prior art, there are many methods for obtaining the positioning tracking parameters of a train, including TDOA, TOA, RSSI, and AOA, etc. Some methods such as TDOA and TOA can directly position the train, but at least three base stations are required to achieve it, while some methods such as AOA do not require the cooperation of multiple base stations, but can only obtain partial positioning parameters and cannot directly position the train.

[0034] The inventor's research found that to achieve single-base station positioning tracking of a train running on a curve, it can be achieved through the cooperation of multiple positioning tracking methods. However, the key to the problem lies in the selection of positioning methods and parameters, as well as the establishment of a suitable mathematical model. In the present invention, the combination of TOA, AOA, and Doppler frequency offset estimation is selected as the combination of train positioning parameters through repeated comparison and experimental screening; secondly, it is necessary to solve the problem of the combination of filtering and positioning methods. The positioning problem of a train on a curve is actually a non-linear system problem, and both the UKF or PF algorithm can well solve the non-linear system problem; at the same time, according to the above positioning parameters and algorithms, the present invention creatively constructs a mathematical model of six-dimensional state variables and five-dimensional observation variables, integrating the state and observation parameters of both the head and the tail into the above mathematical model, which can not only obtain the positioning information of the head and the tail synchronously, but also further improve the estimation accuracy of positioning due to the further increase in the number of parameters of the mathematical model.

[0035] In fact, the present invention also creatively configures multi-antenna arrays at both the front-end and rear-end of the vehicle head, and their functions are as follows: First, through the multi-antenna array, each antenna element is connected to a radio frequency (RF) component, facilitating the use of the direction spectrum method to obtain the angle of arrival (AOA) estimation (the essence of the direction spectrum is to search for the local maximum of the signal space spectrum by analyzing the time series of the received signal to determine the direction of the incoming wave. The MUSIC - Multiple Signal Classification algorithm proposed in 1979 is a classic direction spectrum estimation algorithm. The basic idea is to perform eigenvalue decomposition on the covariance matrix of the received data of any array, thereby obtaining the signal subspace corresponding to the signal components and the noise subspace orthogonal to the signal components, and then using the orthogonality of these two subspaces to estimate the signal parameters); Second, the traditional single-antenna system uses a correlation receiver with the received waveform as the template signal or a matched filter matched to the received waveform, and obtains the time of arrival (TOA) estimation by maximizing the mutual relationship between the received signal and the known template signal. The multi-antenna array system can provide multiple TOA estimation information. By using the multi-antenna time delay estimation method based on the cross-power spectrum to estimate the relative time delay difference of the multi-antenna, and adopting various methods such as signal processing and data fusion, the estimation accuracy of TOA can be significantly improved; Finally, the principle of Doppler frequency shift estimation is to utilize the cyclic prefix characteristic of the signal that overcomes multipath interference, thereby obtaining the Doppler frequency shift estimation by maximizing the similarity between the cyclic prefix and the signal. By using multi-antenna received signals for frequency offset estimation and then taking the average, the performance of Doppler frequency shift estimation can be significantly improved.

[0036] Furthermore, in the measurement equation, by subtracting the measured values of the time of arrival (TOA) at the front and rear of the vehicle head, the clock synchronization error between the base station and the train can be eliminated, the requirement for clock synchronization between the two can be reduced, and the positioning and tracking accuracy of the system can be improved while reducing costs.

[0037] As an optimization, the present invention utilizes the distance information between the front and rear of the vehicle head and the estimation of the state variables, and corrects and updates the position coordinate information of the front and rear of the train head through the least squares algorithm, which can further improve the positioning and tracking accuracy of the train.

[0038] It can be seen that the present invention has the following beneficial effects: By adopting the positioning and tracking method described in the present invention, the feasibility problem of single-base station positioning and tracking of trains on curves is solved, the construction cost is reduced, and the positioning accuracy is further improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] The drawings of the present invention are described as follows.

[0040] Appendix Figure 1 is the layout structure schematic diagram of the present invention.

[0041] Among them, 1. Base station; 2. Head terminal; 3. Tail terminal. Detailed implementation mode

[0042] The present invention will be further described below in conjunction with embodiments.

[0043] As shown in the Figure 1 drawing, the curve where the train travels is covered by the signal of a base station 1 for train tracking. A positioning signal communication terminal is respectively arranged at the head and the tail of the train. The positioning signal communication terminals arranged at the head and the tail are respectively denoted as the head terminal 2 and the tail terminal 3. Both the head terminal 2 and the tail terminal 3 are configured with multi-antenna arrays;

[0044] The positioning and tracking method includes:

[0045] 1) The base station 1 simultaneously sends downlink signals to the head terminal 2 and the tail terminal 3 in real time; the downlink signals include pilot signals and synchronization signals;

[0046] 2) According to the downlink signals obtained by the head terminal 2, three measurement values of the TOA (Time of Arrival), AOA (Angle of Arrival), and Doppler Shift of the head are calculated; at the same time, according to the downlink signals obtained by the tail terminal 3, three measurement values of the TOA, AOA, and Doppler Shift of the tail are calculated;

[0047] 3) The UKF algorithm or the PF algorithm is used to obtain the positioning coordinates of the head at the nth moment and the positioning coordinates of the tail

[0048] In step 3), the state vector equations involved in the UKF algorithm or the PF algorithm are: s[n]=f(s[n - 1]) + q[n], and the measurement equation is: G[n]=h(s[n]) + u[n];

[0049] Among them, n is a certain moment in the discretized sampling time, s[n] is the state variable at the nth moment, and its definition is:

[0050]

[0051] Among them, and are respectively the abscissa and ordinate of the head at the nth moment, v[n] is the train speed at the nth moment, and φ 1 [n] is the running direction angle of the head at the nth moment, and are respectively the abscissa and ordinate of the tail at the nth moment;

[0052] f(s n-1 ) is the non - linear state transition function at time n - 1, and the definition of f(s n-1 ) is as follows:

[0053]

[0054] Among them, s[n - 1] is the state variable at time n - 1; Δt is the sampling time interval; the definition of the said s[n - 1] is:

[0055]

[0056] Among them and are respectively the abscissa and ordinate of the front of the train at time n - 1, v[n - 1] is the train speed at time n - 1 of the train, and φ 1 [n - 1] is the running direction angle of the front of the train at time n - 1, and are respectively the abscissa and ordinate of the rear of the train at time n - 1; φ 2 [n - 1] is the running direction angle of the rear of the train at time n - 1;

[0057] Among them, the relationship between the running direction angle φ 1 [n] of the front of the train and the running direction angle φ 2 [n] of the rear of the train is determined by the following formula:

[0058]

[0059] Among them, L is the formation length of the train, and R is the minimum curve radius of the curve;

[0060] q[n] is the system noise vector at time n, and the said q[n] follows Gaussian white noise with a mean of zero, and the said q[n] is a set value;

[0061] G[n] is the measurement variable at time n, and the definition of G[n] is:

[0062]

[0063] Among them, and are respectively the three measurement values of TOA, AOA and Doppler frequency shift of the front of the train at time n, and are respectively the three measurement values of TOA, AOA and Doppler frequency shift of the rear of the train at time n;

[0064] h(s[n]) is the non - linear measurement function at time n, and the definition of h(s[n]) is:

[0065]

[0066] Among them, B is the position coordinate of the base station, and B = (B x , B y ); c is the speed of light; λ is the wavelength of the signal; the symbol ||||2 represents the second-order norm;

[0067] u[n] is the measurement noise vector at time n, q[n] follows Gaussian white noise with a mean of zero, and q[n] is a set value.

[0068] As an optimization, the communication among the head terminal 2, the tail terminal 3 and the base station 1 adopts a 5G-R network communication system. Compared with the traditional narrowband mobile communication system (GSM-R) and broadband mobile communication system (LTE-R), it can be further improved in terms of mobile communication quality, the number of access devices, and reliable high-data-rate transmission, helping the new generation of railway mobile communication system to achieve further improvement in obtaining satellite-assisted data, real-time monitoring of train operation status, etc.

[0069] As an optimization, the present invention uses the distance information between the head and the tail of the train and the estimation of state variables to correct the position coordinate information of the train head and the train tail through the least squares algorithm, which can further improve the accuracy of train positioning and tracking. Specifically: the positioning coordinates of the train head are updated using Formula 1 to obtain the corrected positioning coordinates of the train head at time n The positioning coordinates of the train tail are updated using Formula 2 to obtain the corrected positioning coordinates of the train tail at time n

[0070] Formula 1 is:

[0071]

[0072] Formula 2 is:

[0073]

[0074] Among them, and are the abscissa and ordinate of the corrected train head at time n, respectively; and are the abscissa and ordinate of the corrected train tail at time n, respectively;

[0075] Among them,

[0076]

[0077] Among them,

[0078] The UKF algorithm, PF algorithm, least squares algorithm, and TOA, AOA, and Doppler frequency offset estimation theories applied in the present invention are all very common processing means or calculation methods in the prior art. For relevant content, those skilled in the art can obtain it from relevant literature of the prior art.

Claims

1. A positioning and tracking method for trains on a curve, characterized in that: The curved track is covered by a base station signal for train tracking. A positioning signal communication terminal is respectively arranged at the head and the tail of the train. The positioning signal communication terminals arranged at the head and the tail are respectively denoted as the head terminal and the tail terminal. Both the head terminal and the tail terminal are configured with multi-antenna arrays; The positioning and tracking method includes: 1) The base station simultaneously sends downlink signals to the head terminal and the tail terminal in real time. The downlink signals include pilot signals and synchronization signals; 2) Calculate three measurement values of TOA, AOA, and Doppler frequency shift of the train head according to the downlink signals obtained by the head terminal. At the same time, calculate three measurement values of TOA, AOA, and Doppler frequency shift of the train tail according to the downlink signals obtained by the tail terminal; (III) Obtain the positioning coordinates of the vehicle head at the nth moment by using the UKF algorithm or the PF algorithm and the positioning coordinates of the vehicle tail In the step 3), the state vector equations involved in the UKF algorithm or PF algorithm are: s[n] = f(s[n - 1]) + q[n], and the measurement equation is: G[n] = h(s[n]) + u[n]; where n is a certain moment in the discretized sampling time, s[n] is the state variable at the nth moment, and its definition is: Among them, and are the abscissa and ordinate of the train head at time n respectively, v[n] is the train speed at time n, and φ 1 [n] is the running direction angle of the train head at time n, and are the abscissa and ordinate of the train tail at time n respectively; f(s n-1 ) is the non-linear state transition function at time n - 1, and the definition of f(s n-1 ) is as follows: where s[n - 1] is the state variable at the (n - 1)th moment; Δt is the sampling time interval; the definition of s[n - 1] is: Among them and are respectively the abscissa and ordinate of the train head at the (n - 1)th moment, v[n - 1] is the train speed at the (n - 1)th moment, and φ 1 [n - 1] is the running direction angle of the train head at the (n - 1)th moment, and are respectively the abscissa and ordinate of the train tail at the (n - 1)th moment; φ 2 [n - 1] is the running direction angle of the train tail at the (n - 1)th moment; Among them, the running direction angle φ 1 [n] of the vehicle head and the running direction angle φ 2 [n] of the vehicle tail are determined by the following formula: where L is the formation length of the train and R is the minimum curve radius of the curved track; q[n] is the system noise vector at the nth moment. The q[n] follows Gaussian white noise with a mean of zero, and the q[n] is a set value; G[n] is the measurement variable at the nth moment, and the definition of G[n] is: Among them, and are respectively the three measurement values of TOA, AOA and Doppler frequency shift at the front of the vehicle at time n, and are respectively the three measurement values of TOA, AOA and Doppler frequency shift at the rear of the vehicle at time n; h(s[n]) is the non-linear measurement function at the nth moment, and the definition of h(s[n]) is: Among them, B is the position coordinate of the base station, and the B = (B x , B y ); c is the speed of light; λ is the wavelength of the signal; the symbol ‖‖2 represents the second-order norm; u[n] is the measurement noise vector at the nth moment. The q[n] follows Gaussian white noise with a mean of zero, and the q[n] is a set value.

2. The positioning and tracking method of a train on a curve according to claim 1, characterized in that: The communication among the head terminal, the tail terminal, and the base station adopts a 5G-R network communication system.

3. The method for positioning and tracking a train on a curve according to claim 1 or 2, characterized in that: Update the T 1 [n] in step (iii) using the least squares algorithm to obtain the corrected head positioning coordinates at time n where and are the abscissa and ordinate of the corrected head at time n respectively; Update the T 2 [n] in step (iii) using the least squares algorithm to obtain the corrected tail positioning coordinates at time n where and are the abscissa and ordinate of the corrected tail at time n respectively.

Images