CN116753958A

CN116753958A - Train positioning method, device, equipment and medium based on Kalman filtering algorithm

Info

Publication number: CN116753958A
Application number: CN202310728567.XA
Authority: CN
Inventors: 樊宽刚; 胡倩; 徐艺玮; 肖伟兵
Original assignee: Jiangxi University of Science and Technology
Current assignee: Jiangxi University of Science and Technology
Priority date: 2023-06-19
Filing date: 2023-06-19
Publication date: 2023-09-15

Abstract

The application provides a train positioning method, a device, equipment and a medium based on a Kalman filtering algorithm, wherein the first position information of a target train in an SINS inertial navigation system, the second position information of the target train in a GNSS global navigation satellite system and the third position information of the target train acquired by Tag labels arranged on the target train are respectively acquired; determining initial position observation values of the target trains in the positioning subsystems according to the first position information, the second position information and the third position information; determining the observability degree of each positioning subsystem according to the state vector of the initial position observation value and the observation noise; determining candidate position observation values of the target trains according to the observability of each positioning subsystem; and filtering the candidate position observation value of the target train in the train positioning system by using an unscented Kalman filtering algorithm to obtain target position information. By adopting the method, the high-precision positioning of the train is realized.

Description

Train positioning method, device, equipment and medium based on Kalman filtering algorithm

Technical Field

The application relates to the technical field of vehicle positioning, in particular to a train positioning method, device, equipment and medium based on a Kalman filtering algorithm.

Background

Along with the development of magnetic levitation technology, a suspension type permanent magnet magnetic levitation railway transportation system is mature, and the permanent magnet magnetic levitation railway train has the remarkable characteristics of green energy conservation, intelligent driving, safety, reliability, economy and practicability. Because the magnetic levitation train is easily influenced by various factors in operation, the position and the operation condition of the magnetic levitation train are known in real time, the safety protection of the magnetic levitation train is realized, and the key problem is the accurate positioning of the magnetic levitation train.

In the prior art, a single sensor or a single positioning system is typically used to position the train. However, the suspension type permanent magnetic levitation track traffic system has the problems of more control variables, low automation degree, high integration level of the control system, high hardware complexity, strict requirement on real-time control and the like, for example, when a permanent magnetic levitation train passes through a special section such as a big tree shelter or a tunnel, part of sensors are interfered, or a single system breaks down and cannot acquire effective data, so that the problems of poor robustness, low positioning precision, poor stability and the like of sensor information fusion caused by adopting the single system single sensor to position the single system single sensor are caused, and the high-precision positioning of the train is difficult to realize.

Disclosure of Invention

In view of the above, the present application aims to provide a method, a device, equipment and a medium for positioning a train based on a kalman filter algorithm, so as to realize high-precision positioning of the train.

In a first aspect, an embodiment of the present application provides a method for positioning a train based on a kalman filter algorithm, where the method includes:

respectively acquiring first position information of a target train in an SINS inertial navigation system at each moment, second position information of the target train in a GNSS global navigation satellite system and third position information of the target train acquired by Tag tags arranged on the target train;

determining a first initial position observation value of the target train at each moment in a first positioning subsystem by using a Kalman filtering algorithm according to the first position information and the second position information at each moment, wherein the first positioning subsystem consists of the inertial navigation system and the global navigation satellite system;

determining a second initial position observation value of the target train at each moment in a second positioning subsystem by using a Kalman filtering algorithm according to the first position information and the third position information at each moment, wherein the second positioning subsystem consists of the inertial navigation system and the Tag label;

Determining a first observable degree of the first positioning subsystem at each time according to a state vector of a first initial position observed value of the target train at each time and observation noise of a first initial position observed value of the target train at each time, and determining a second observable degree of the second positioning subsystem at each time according to a state vector of a second initial position observed value of the target train at each time and observation noise of a second initial position observed value of the target train at each time;

determining a candidate position observation value of the target train in a train positioning system at each moment according to a first observability degree of the first positioning subsystem at each moment and a second observability degree of the second positioning subsystem at each moment, wherein the train positioning system consists of the inertial navigation system, the global navigation satellite system and the Tag;

and respectively filtering candidate position observation values of the target train at each time point in the train positioning system by using an unscented Kalman filtering algorithm to obtain target position information of the target train at each time point.

Optionally, the determining the candidate position observation value of the target train in the train positioning system according to the first observability degree of the first positioning subsystem at each time and the second observability degree of the second positioning subsystem at each time comprises:

determining a first target position observation value of the first positioning subsystem at each time according to a first observability value of the first positioning subsystem at each time and a first initial position observation value of the target train at each time;

determining a second target position observation value of the second positioning subsystem at each time according to a second observability value of the second positioning subsystem at each time and a second initial position observation value of the target train at each time;

and determining a candidate position observation value of the target train at each time according to the first target position observation value of the first positioning subsystem at each time and the second target position observation value of the second positioning subsystem at each time.

Optionally, the determining the first target position observed value of the first positioning subsystem at each time according to the first observable value of the first positioning subsystem at each time and the first initial position observed value of the target train at each time includes;

For each moment, determining the product of a first observability value of the first positioning subsystem at the moment and a first initial position observation value of the target train at the moment as a first target position observation value of the first positioning subsystem at the moment;

the determining, according to the second observable degree of the second positioning subsystem at each time and the second initial position observed value of the target train at each time, the second target position observed value of the second positioning subsystem at each time includes:

for each moment, determining the product of a second observability value of the second positioning subsystem at the moment and a second initial position observation value of the target train at the moment as a second target position observation value of the second positioning subsystem at the moment;

the determining the candidate position observation value of the target train at each time according to the first target position observation value of the first positioning subsystem at each time and the second target position observation value of the second positioning subsystem at each time comprises the following steps:

and for each moment, determining the sum of the first target position observed value of the first positioning subsystem at the moment and the second target position observed value of the second positioning subsystem at the moment as a candidate position observed value of the target train at each moment.

Optionally, after filtering candidate position observations of the target train at each moment in the train positioning system by using an unscented kalman filtering algorithm to obtain target position information of the target train at each moment, the method further includes;

and generating the running track of the target train according to the target position information of the target train at each moment and the time information of each moment.

In a second aspect, an embodiment of the present application provides a train positioning device based on a kalman filtering algorithm, where the device includes:

the position information determining module is used for respectively acquiring first position information of the target train in the SINS inertial navigation system at each moment, second position information of the target train in the GNSS global navigation satellite system and third position information of the target train acquired by Tag labels arranged on the target train;

the first initial position observation value determining module is used for determining a first initial position observation value of the target train at each moment in a first positioning subsystem according to the first position information and the second position information at each moment by using a Kalman filtering algorithm, wherein the first positioning subsystem consists of the inertial navigation system and the global navigation satellite system;

The second initial position observation value determining module is used for determining a second initial position observation value of the target train at each moment in a second positioning subsystem according to the first position information and the third position information at each moment by using a Kalman filtering algorithm, wherein the second positioning subsystem consists of the inertial navigation system and the Tag;

the observability determining module is used for determining a first observability degree of the first positioning subsystem at each moment according to a state vector of a first initial position observed value of the target train at each moment and observation noise of the first initial position observed value of the target train at each moment, and determining a second observability degree of the second positioning subsystem at each moment according to a state vector of a second initial position observed value of the target train at each moment and observation noise of the second initial position observed value of the target train at each moment;

the candidate position observation value determining module is used for determining a candidate position observation value of the target train in the train positioning system at each moment according to a first observability degree of the first positioning subsystem at each moment and a second observability degree of the second positioning subsystem at each moment, wherein the train positioning system consists of the inertial navigation system, the global navigation satellite system and the Tag;

And the target position information determining module is used for respectively filtering candidate position observation values of the target train at each time in the train positioning system by using an unscented Kalman filtering algorithm to obtain target position information of the target train at each time.

Optionally, the candidate position observation value determining module is configured to, when determining the candidate position observation value of the target train in the train positioning system at each time according to the first observability degree of the first positioning subsystem at each time and the second observability degree of the second positioning subsystem at each time, specifically:

Optionally, the candidate position observation value determining module is configured to, when determining the first target position observation value of the first positioning subsystem at each time point according to the first observable value of the first positioning subsystem at each time point and the first initial position observation value of the target train at each time point, specifically:

the candidate position observation value determining module is configured to, when determining, according to a second observable value of the second positioning subsystem at each time and a second initial position observation value of the target train at each time, a second target position observation value of the second positioning subsystem at each time, specifically:

The candidate position observation value determining module is specifically configured to, when determining the candidate position observation value of the target train at each time according to the first target position observation value of the first positioning subsystem at each time and the second target position observation value of the second positioning subsystem at each time:

Optionally, the apparatus further comprises:

and the running track determining module is used for respectively filtering candidate position observation values of the target train at each moment in the train positioning system by using a unscented Kalman filtering algorithm to obtain target position information of the target train at each moment, and then generating the running track of the target train according to the target position information of the target train at each moment and the time information of each moment.

In a third aspect, an embodiment of the present application provides a computer apparatus, including: a processor, a memory and a bus, the memory storing machine readable instructions executable by the processor, the processor and the memory in communication via the bus when the computer device is running, the machine readable instructions when executed by the processor performing the steps of the kalman filter algorithm based train positioning method of any of the alternative embodiments of the second aspect described above.

In a fourth aspect, embodiments of the present application provide a computer readable storage medium having stored thereon a computer program which, when executed by a processor, performs the steps of the kalman filter algorithm based train positioning method described in any of the alternative embodiments of the second aspect above.

The technical scheme provided by the application comprises the following beneficial effects:

respectively acquiring first position information of a target train in an SINS inertial navigation system at each moment, second position information of the target train in a GNSS global navigation satellite system and third position information of the target train acquired by Tag tags arranged on the target train; through the steps, the position information of the target train acquired by a plurality of different positioning systems can be acquired.

Determining a first initial position observation value of the target train at each moment in a first positioning subsystem by using a Kalman filtering algorithm according to the first position information and the second position information at each moment, wherein the first positioning subsystem consists of the inertial navigation system and the global navigation satellite system; determining a second initial position observation value of the target train at each moment in a second positioning subsystem by using a Kalman filtering algorithm according to the first position information and the third position information at each moment, wherein the second positioning subsystem consists of the inertial navigation system and the Tag label; through the steps, a plurality of positioning systems can be combined to obtain different positioning subsystems, and fusion position observation values of the target train under each positioning subsystem are determined, so that fusion of position information of the first target train is realized.

Determining a first observable degree of the first positioning subsystem at each time according to a state vector of a first initial position observed value of the target train at each time and observation noise of a first initial position observed value of the target train at each time, and determining a second observable degree of the second positioning subsystem at each time according to a state vector of a second initial position observed value of the target train at each time and observation noise of a second initial position observed value of the target train at each time; through the steps, the observability of the target train in each positioning subsystem can be obtained.

Determining a candidate position observation value of the target train in a train positioning system at each moment according to a first observability degree of the first positioning subsystem at each moment and a second observability degree of the second positioning subsystem at each moment, wherein the train positioning system consists of the inertial navigation system, the global navigation satellite system and the Tag; through the steps, the position observation values of the target trains under each positioning subsystem can be fused to obtain candidate position observation values of the target trains under the total positioning subsystem, so that the fusion of the position information of the target trains for the second time is realized.

Respectively filtering candidate position observation values of the target train at each time point in the train positioning system by using an unscented Kalman filtering algorithm to obtain target position information of the target train at each time point; through the steps, the position information of the target train obtained after fusion can be filtered to obtain more accurate target position information.

By adopting the method, the position information of the target trains in the systems is acquired respectively, the position information of each two systems is fused by utilizing a Kalman filtering algorithm to obtain the position information of the target trains under the subsystems formed by each two systems, the position information of the target trains under each subsystem is fused again according to the observable degree of each subsystem to obtain the position information of the target trains under the whole positioning system, and finally the position information of the target trains under the whole positioning system is filtered to obtain the more accurate position information of the target trains, so that the high-precision positioning of the trains is realized.

In order to make the above objects, features and advantages of the present invention more comprehensible, preferred embodiments accompanied with figures are described in detail below.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings that are needed in the embodiments will be briefly described below, it being understood that the following drawings only illustrate some embodiments of the present invention and therefore should not be considered as limiting the scope, and other related drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.

Fig. 1 shows a flowchart of a train positioning method based on a kalman filter algorithm according to an embodiment of the present invention;

FIG. 2 is a flow chart of a candidate position observation determination method according to a first embodiment of the invention;

fig. 3 shows a schematic structural diagram of a train positioning device based on a kalman filter algorithm according to a second embodiment of the present invention;

fig. 4 shows a schematic structural diagram of a second train positioning device based on a kalman filtering algorithm according to a second embodiment of the present invention;

fig. 5 shows a schematic structural diagram of a computer device according to a third embodiment of the present invention.

Detailed Description

For the purpose of making the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention, and it is apparent that the described embodiments are only some embodiments of the present invention, not all embodiments. The components of the embodiments of the present invention generally described and illustrated in the figures herein may be arranged and designed in a wide variety of different configurations. Thus, the following detailed description of the embodiments of the invention, as presented in the figures, is not intended to limit the scope of the invention, as claimed, but is merely representative of selected embodiments of the invention. All other embodiments, which can be made by a person skilled in the art without making any inventive effort, are intended to be within the scope of the present invention.

Example 1

For the convenience of understanding the present application, the following describes the first embodiment of the present application in detail with reference to the flowchart of the first embodiment of the present application provided by the kalman filter algorithm-based train positioning method.

Referring to fig. 1, fig. 1 shows a flowchart of a method for positioning a train based on a kalman filter algorithm according to an embodiment of the present application, where the method includes steps S101 to S106:

s101: and respectively acquiring first position information of a target train in the SINS inertial navigation system at each moment, second position information of the target train in the GNSS global navigation satellite system and third position information of the target train acquired by Tag tags arranged on the target train.

Specifically, the INS inertial navigation system is based on Newton mechanical inertia principle, adopts an accelerometer and a gyroscope to acquire acceleration information and angular velocity information of a carrier, and obtains position information of the carrier through twice integration operation of the acceleration, so that the INS inertial navigation system is a navigation system with high accuracy. The INS is mainly divided into a GINS platform type inertial navigation system consisting of an entity inertial platform and a strapdown SINS inertial navigation system consisting of a mathematical calculation platform according to the difference of structural principles. SINS can realize all-weather continuous three-dimensional orientation and positioning work in the global scope by the system without depending on external information, has strong anti-interference capability, does not need to exchange data with the outside of a carrier, and has the characteristics of strong autonomy and good concealment.

The GNSS global navigation satellite system is based on artificial positioning satellites, and is a global navigation positioning system based on a wireless communication technology. The GNSS can obtain the longitude and latitude of the target, the position information of the target can be obtained through iterative calculation of a longitude and latitude calculation formula between two points, radio signals with different frequencies can be continuously transmitted, the GPS can provide heading, elevation, speed and other vehicle information for a train in 24 hours in an all-weather mode when being cooperated with a ground receiving station, the acquired measurement data are updated in real time, and the GPS positioning device has the advantages of being high in positioning precision and capable of working at all times.

Tag tags are mainly used to detect the accuracy of train positioning. The Tag is arranged on two sides of a train, the card reader is arranged on the track, when the train runs through the track with the Tag card reader in the running process of the train, the card reader reads out data in the Tag, after data filtering and data conversion, real-time data of the train are sent to the information fusion control center for detecting the accuracy degree of the information fusion control center on GNSS positioning data and SINS positioning data, and because the positioning data of the Tag of the train are accurate in determination, the ground information fusion control center can correct the positioning data of the train fused by the positioning means according to the positioning data of the Tag of the train so as to improve the positioning accuracy of the permanent magnetic levitation train.

S102: and determining a first initial position observation value of the target train at each moment in a first positioning subsystem by using a Kalman filtering algorithm according to the first position information and the second position information at each moment, wherein the first positioning subsystem consists of the inertial navigation system and the global navigation satellite system.

S103: and determining a second initial position observation value of the target train at each time point in a second positioning subsystem by using a Kalman filtering algorithm according to the first position information and the third position information at each time point, wherein the second positioning subsystem consists of the inertial navigation system and the Tag label.

Specifically, the invention selects a sensor such as SINS, GNSS, tag Tag, wherein SINS is used as a basic navigation system, GNSS and Tag are used as external auxiliary navigation systems, and the SINS/GNSS and SINS/Tag sub-systems are respectively combined into two sub-systems, wherein the SINS/GNSS sub-system is the first positioning sub-system, and the SINS/Tag sub-system is the second sub-system.

Multiple sensors provide redundant information when individual sensors in a multiple sensor system measure the same characteristic parameter of a target with different confidence levels. Because the observation noise of single sensor in the multi-sensor system is not related to each other, the redundant information obtained by measurement can be fused, so as to reduce the uncertainty of the system and improve the precision and reliability of the system. In addition, due to the presence of redundant information, when one or more sensors fail, the system can still maintain normal operation using information acquired by other sensors.

The SINS/GNSS/Tag label combined positioning system of the maglev train adopts a two-layer structure form: firstly, SINS in a first layer is respectively combined with GNSS and Tag tags to form two sub-filters working simultaneously, and corresponding optimal estimation results are obtained through synchronous processing of measurement data; the data is further processed at the second layer by a main filter. When the satellite signals of the magnetic suspension train positioning system lose lock, the SINS and the Tag can work independently, and the measurement information of the Tag is used for correcting the positioning output of the SINS; when the satellite signal is unstable, the SINS, the GNSS and the Tag label work simultaneously, and the Tag label is adopted to correct the output of the SINS/GNSS.

The design of the kinetic model is based on the design of a single particle column model. Maglev trains are considered to be one particle, and interactions between vehicles are not considered. The equation of motion of the dynamic maglev train model is as follows.

v _k ＝v _k-1 +a×Δk (2)

Wherein x is _k And x _k-1 Respectively represent the positions of the magnetic levitation trains at the time k and the time k-1, v _k And v _k-1 The speeds of the magnetic levitation train at the time k and the time k-1 are respectively represented, Δk is the time interval, and a is the acceleration of the magnetic levitation train.

For SINS/GNSS integrated navigation system, SINS/GNSS measurement vector Respectively consisting of the difference between the position coordinates and velocity calculated by the SINS and the corresponding values measured by the GNSS. So the error measurement equation of the SINS/GNSS integrated navigation system is as follows:

wherein, the liquid crystal display device comprises a liquid crystal display device,vector measurement for SINS/GNSS subsystem, < >>To observe the matrix, x _k+1 Is SINS/GNSS subsystem state vector, < ->To observe noise.

Similarly, the error measurement equation of the SINS/Tag label integrated navigation system is as follows:

wherein, the liquid crystal display device comprises a liquid crystal display device,vector measurement for SINS/Tag subsystem, < >>To observe the matrix, x _k+1 Is SINS/Tag subsystem state vector, < >>To observe noise.

The navigation error model, the error state and the measurement equation play a vital role in the processing and the fusion of navigation information in the proposed self-adaptive unscented Kalman filtering.

After the steps are finished, redundant information preprocessing is carried out on data measured by multiple types of sensors, the information quality is improved by utilizing the complementarity of multi-source information, the observability (DOO) of each combined subsystem is calculated, the information quality and the working performance of the subsystems are evaluated according to the size of DOO values, weight distribution is carried out on observed values, a magnetic levitation train mathematical model based on multi-sensor information fusion positioning is established, adaptive filtering is used for information fusion, in the whole information fusion process, when the observed noise covariance matrix is unmatched with noise appearing in the current filtering process, the observed noise covariance matrix is corrected in real time by adopting a covariance matching technology, finally, the position information of the permanent magnetic levitation train is obtained, the multi-source information correlation of the permanent magnetic levitation train positioning is revealed, and the spatial positioning precision of the train is further improved.

S104: determining a first observable degree of the first positioning subsystem at each time according to a state vector of a first initial position observed value of the target train at each time and observation noise of a first initial position observed value of the target train at each time, and determining a second observable degree of the second positioning subsystem at each time according to a state vector of a second initial position observed value of the target train at each time and observation noise of a second initial position observed value of the target train at each time.

In particular, the observability is an important concept for error state estimation in integrated navigation, as it determines the existence and nature of the navigation solution. Typically, the observability analysis of a linear time-invariant system is simple, whereas the analysis of a linear time-variant system is cumbersome and requires an assessment of the observability Gramian. The Gramian criterion can only give a "yes-no" type of answer and cannot account for the relative observability of all state variables. In order to describe the observable ability of each navigation error more accurately, analysis based on DOO values is highly desirable.

For integrated navigation systems, the navigation error model of each subsystem has a great impact on the accuracy and performance of the overall system. Due to the variety of sensor types, the error models of integrated navigation systems are combined into one model complex, where SINS, GNSS, tag label error models exist, etc.

Let the discrete form of the error state be expressed as:

wherein X is _k Represents the position state value of the train at the moment k, f (·) represents the nonlinear state transfer function, X _k-1 Representing the position state value, w, of the train at time k-1 _k-1 Representing process noise, Z _k Represents the position observation value of the train at the moment k, H (x) represents a nonlinear measurement function, and H _k Observation matrix v representing k time _k Representing observed noise.

The discrete form of error state according to (5) can introduce a new measurement equation:

wherein, the liquid crystal display device comprises a liquid crystal display device,o is the observed value of the sensor _k Representing an observation matrix, X _k A state vector representing the sensor at time k, < >>Representing observed noise.

Considering the pseudo-inverse of the observability matrix, we can get the relationship between the systematic error state and the measurement:

where Θ represents the pseudo-inverse of the matrix,representing the observation matrix O _k Is a pseudo-inverse of (a).

Assume thatThen it is expressed in scalar form:

wherein, the liquid crystal display device comprises a liquid crystal display device,representing vector Γ _k I-th component of>Representation of ψ _k I-th component of>Representation matrix->J=1, 2, … n, n is the number of state variables, +.>For the observed quantity of the kth sensor,for the observation of the k+1th sensor, -/->For the observation of the k+n-1 th sensor, < > >Observation noise for kth sensor, < ->Observation noise for the (k+1) th sensor,>is the observation noise of the k+n-1 th sensor.

Thus, a linear time-varying systemVector form of observability measure of systemThe calculation is as follows:

wherein, the liquid crystal display device comprises a liquid crystal display device,

wherein E is _k [(X ⁱ ) ² ]Representing the ith state variable variance, E _k [(Γ ⁱ ) ² ]Represents the variance of the measured values, k represents the time, l represents the number of state variables, X ⁱ Represents the ith state variable, E _k The standard deviation is represented by the term "standard deviation",is the ith state variable at the jth moment, < >>Is the ith measurement at the jth time.

Conversion to scalar formThe following are provided:

wherein e _k Representing standard deviation, x ⁱ Representing the i-th state variable value, E _k [(x ⁱ ) ² ]Representing the variance of the ith state variable, E _k [(Γ ⁱ ) ² ]Representing the variance of the measured value,representing the j-th state variable in the matrix +.>N represents the number of state variables, +.>For the observable degree of the ith positioning subsystem at time k, when i=1,/is +.>The method comprises the steps that the first positioning subsystem is the observable degree of the first positioning subsystem at the k moment, namely the first observable degree at the k moment; when i=2, _a->Is the observable of the second positioning subsystem at the k moment, namely the second observable at the k moment.

S105: and determining a candidate position observation value of the target train in a train positioning system at each moment according to the first observability of the first positioning subsystem at each moment and the second observability of the second positioning subsystem at each moment, wherein the train positioning system consists of the inertial navigation system, the global navigation satellite system and the Tag.

In particular, in integrated navigation systems, the performance of each sensor may be different due to changes in the external environment, in which case some sensors operate in good conditions, while others operate in harsh conditions. Therefore, it is necessary to evaluate the operation performance of each sensor. In the invention, a scalar method is proposed to measure the observability of a linear time-varying system. Based on DOO criteria, a new sensor performance numerical evaluation method is provided. The degree of observability can help us obtain the observability degree information of the state variables and can further serve as an index for evaluating the performance of each navigation sensor and autonomously managing the sensors in the integrated navigation system. Thus, the new algorithm may help provide a relatively high precision navigation solution in a continuous manner. The invention is beneficial to the development of observability theory and provides digital indexes for optimizing control and navigation system design for developers. And distributing corresponding weights to the observed values of each subsystem according to the calculated DOO values, so as to carry out filtering fusion on the observed values distributed with the weights.

Due toTherefore, the DOO value calculated by each subsystem can be used as the weight of the observed value to obtain a new train position observed value:

Wherein, the liquid crystal display device comprises a liquid crystal display device,for the observable amount of the ith positioning subsystem at time k, when i=1,/is>The method comprises the steps that the first positioning subsystem is the observable degree of the first positioning subsystem at the k moment, namely the first observable degree at the k moment; when i=2, _a->For the second positioning subsystem the observability at time k, i.e. the second observability at time k,/for the second positioning subsystem>Is a candidate position observation at time k, < ->And the measurement vector represents the train position of the ith positioning subsystem at the moment k.

S106: and respectively filtering candidate position observation values of the target train at each time point in the train positioning system by using an unscented Kalman filtering algorithm to obtain target position information of the target train at each time point.

Specifically, the content of the standard unscented kalman filter algorithm is as follows:

considering a general nonlinear discrete time dynamic system, the process and measurement model can be described as follows:

wherein X is _k ∈R ^n×1 Represents the train position state vector at time k, Z _k ∈R ^m×1 Representing train position observation measurement vectors at time k, f (·) and H (·) representing known nonlinear state transfer and measurement functions, respectively, H _k Representing an observation matrix, X _k Represents the train position state vector at time k, X _k-1 Representing a train position state vector at time k-1, w _k-1 V is process noise _k To observe noise.

Let w be _k-1 And v _k Is uncorrelated zero-mean Gaussian white noise, w _k-1 Process noise, v, being zero-mean gaussian white noise _k Observation noise, w, being zero-mean gaussian white noise _k-1 Covariance of Q _k-1 ，v _k Covariance of R _k The method comprises the following steps:

wherein, the liquid crystal display device comprises a liquid crystal display device,for process noise covariance, ++>For measuring noise covariance +.>Is the cross covariance of the process noise and the observation noise, w _k V is process noise _k Measurement noise, Q _k Is a process noise covariance and is not a negative definite matrix, R _k Is the measured noise covariance and is a positive definite matrix.

The UKF algorithm is based on the concept that it should be easier to estimate a nonlinear distribution than to approximate a nonlinear function. In the standard UKF, an unscented transformation is implemented to generate sigma points to be non-linearly transformed and the first two matrices of the transformation set are calculated.

The general procedure is as follows:

step one: initializing:

wherein, the liquid crystal display device comprises a liquid crystal display device,is the initial average value, E { X ] ₀ The } represents the initial mean, X ₀ Is the initial value, T is the transposed matrix symbol, P ₀ Is the initial error covariance.

Step two: sigma point calculation:

wherein, the liquid crystal display device comprises a liquid crystal display device,for the first sigma point mean at time k-1,/>Predicting the mean value for the moment k-1, +.>Mean value of ith sigma point at k-1 moment, P _k-1 Representing the error covariance at time k-1, n is the dimension of the state, λ is the complex scale factor, and:

λ＝α ² (n+k)-n (18)

where a is a first adjustment parameter, κ is a second adjustment parameter, the parameter α is typically set to 0.ltoreq.α.ltoreq.1, and k is typically set to 0 by default.

Step three: state prediction:

P _k ＝P _XX +Q _k-1 (22)

wherein, the liquid crystal display device comprises a liquid crystal display device,indicating the time kThe predictive mean value of the ith sigma point, f is a known nonlinear state transfer function,representing the predicted measure of the ith sigma point at time k-1, _k representing the prediction error covariance at time k, P _XX For state measurement covariance, Q _k-1 Representing process noise->And->Weight vectors, mean and covariance respectively, n being the dimension of the state, +.>The predicted mean at time k is defined as:

wherein the subscript indicates which sample point, m indicates the mean, c indicates the covariance,as the weight of the initial average value,for initial covariance weights, for gaussian distribution, the optimal setting of β is β=2, λ is the complex scale factor, n is the dimension of the state, and α is the first tuning parameter.

Step four: measurement and prediction:

wherein, the liquid crystal display device comprises a liquid crystal display device,representing a measurement prediction of the ith sigma point at time k, a nonlinear measurement function of known h (·),represents the predicted mean value of the ith sigma point at time k, k represents time,/ >Is a measurement of the train position,/->Is composed of Z _k The new train position measurement is multiplied by the weight.

Step five: kalman gain calculation:

/>

wherein P is _XZ Measuring cross covariance for state, P _ZZ To assist in new informationVariance, K _k For Kalman gain, R _k To observe noise covariance.

Step six: and (5) filtering and updating:

wherein, the liquid crystal display device comprises a liquid crystal display device,for posterior estimation, P _k+1 Representing update error covariance,/>For K _k Is a transposed matrix of (a).

And step seven, repeating the steps 2-6 until all sample points are calculated.

In the above standard filtering algorithm, the noise covariance matrix needs to be set to an initial value during initialization, but as the iteration number increases, if the noise covariance matrix is not matched with the noise occurring in the current filtering process, the standard UKF algorithm does not correct the noise covariance matrix in real time, so that the algorithm lacks applicability to time-varying noise, resulting in a final estimation result error becoming larger and even the algorithm not converging, and therefore, the real-time adaptive correction of the noise covariance matrix is necessary. If the noise covariance is calculated at each moment, the algorithm is more complex, the calculated amount is increased, and the final navigation positioning accuracy is affected. In summary, only when the filtering is abnormal, the noise covariance is corrected, so that the workload can be greatly reduced, and the positioning accuracy is improved.

The improved unscented Kalman filtering algorithm used in the invention specifically comprises the following steps:

step one: initializing the same as the standard UKF:

Step two: the sigma point is calculated according to equation (17):

step three: judging whether filtering is abnormal or not:

determining target state observations obtained at a sensor at time k using covariance matchingThere is or is not a mismatch in the filter states. Thus, the new information concept can be introduced, and the new information is assumed to be zeta _k Representing real-time estimated error information, sum of squares of innovationThe innovation theory covariance matrix M is:

wherein H is _k In order to observe the matrix,represents the covariance at time k,/>Representing the transpose of the observation matrix,/, of>Representing observed noise covariance.

In the standard unscented Kalman filtering, the information form of the Kalman filtering is adopted, and the difference between the actual measured value and the predicted value is defined as an innovation sequence:

wherein Z is _k Is a measurement of the train position, H _k In order to observe the matrix,representing the train position state vector at time k.

Defining the difference between the actual measurement value and the estimated value as a residual sequence eta _k ：

Wherein, the liquid crystal display device comprises a liquid crystal display device,representing a train position status vector at time k + 1.

The basis for judging that the filtering is abnormal is as follows:

wherein, the liquid crystal display device comprises a liquid crystal display device,is the sum of the squares of the innovation, H _k For observing matrix +.>For the covariance at time k +.>For transpose of the observation matrix +.>Observed noise covariance。

If equation (37) does not hold, indicating that no anomaly has occurred in the current filtering,turning the algorithm to a step five for the self-adaptive observed noise covariance; otherwise, it is indicated that the filtering is abnormal, the observed noise covariance matrix is not matched with the current filtering, the current observed data should be emphasized more, the current observed noise covariance matrix needs to be updated to adapt to the current filtering process, and the algorithm goes to the step four.

Step four: correcting the observed noise covariance matrix in real time:

substituting formula (12) into (35) yields:

definition:

wherein ζ _k To be new and new, Z _k A measurement vector representing the train position at time k,a predictive state vector representing the train position at time k, H _k To observe the matrix, X _k V is the true state vector of the train position at time k _k To observe noise, X _k+1 Represents the real state vector of the train position at time k+1, -/-, for example>Represents the train position prediction state vector at time k +1,is the train position state vector offset value at time k+1.

So that from the definition of the prediction covariance and the error covariance it is possible to:

wherein, the liquid crystal display device comprises a liquid crystal display device,representing the prediction error covariance at time k, +.>Representing the prediction error covariance at time k +1,representing prediction error covariance,/>Representing prediction error covariance,/>Is->Transposed matrix of>Is->Transposed matrix of>For the train position state vector at time kDeviation value.

Due to orthogonality between observed noise and state estimation errors, the innovation covariance can be calculated from (35) as:

wherein, the liquid crystal display device comprises a liquid crystal display device,is the sum of the squares of the innovation->For observing noise covariance +.>Is the innovation covariance, P _k The prediction error covariance at time k is indicated.

A recursive estimation formula is used considering the smoothness of the covariance estimation. Finally, the measurement noise covariance is calculated as follows:

q _k ＝(1-b)(1-b ^k+1 ) ^-1 (46)

wherein q _k In order to adjust the proportion of noise in real time according to the system and the external condition, thereby enhancing the reliability and the credibility of the system, reducing the noise and the external interference of the system, improving the capability of coping with abrupt changes of the state and enhancing the tracking performance. b is a genetic element and 0<b<1, the size of the genetic factor can change the memory length of the filter. Proper value of b is an important guarantee of error convergence of the filtering algorithm, and if the noise covariance changes rapidly, more importance should be attached The influence b of the current observation data on the estimation performance at the moment should take a larger value;to adaptively observe the noise covariance, H _k To observe the matrix, P _k Error covariance at time k +.>For transpose of the observation matrix ε _k For innovation, k is time.

The optimal forgetting factor b is as follows:

wherein, the liquid crystal display device comprises a liquid crystal display device,represents a predicted measurement vector of the train position at time k, tr represents a trace, < >>Transpose of observation matrix, Q _k For process noise covariance, A is the state transition matrix, H _k For observing matrix +.>For the covariance matrix at time k +.>Is the observed noise covariance after adaptation.

Step five: from equation (27):

wherein, the liquid crystal display device comprises a liquid crystal display device,representing predicted observations of train position,/->Represents the mean weight>Total observation value representing the train position of two subsystems at time k,/>DOO value representing subsystem, +.>A measurement vector representing the train position at time k.

Step six: calculating Kalman gain and filtering update;

K _k ^′ ＝P _X ^′ _Z (P _Z ^′ _Z ) ^-1 (51)

P _k ^′ ₊₁ ＝P _k -K _k ^′ P _Z ^′ _z (K _k ^′ ) ^T (53)

wherein P is _X ^′ _Z Representing state measurement cross covariance, P _Z ^′ _Z Representing the covariance of the innovation in the form of,representing posterior estimates, P _k ^′ ₊₁ Error covariance at time k+1, < ->Is the mean weight>Representing the predicted mean value of the ith sigma point at time k,/>Measurement prediction representing the ith sigma point at time k,/i >Representing predicted observations of train position,/->For adaptive observation noise covariance, +.>Representing the total observed value of train positions of two subsystems at time K _k ^′ Representing Kalman gain, P _k The error covariance at time k is indicated.

7. Repeating the steps 2-6 until the end.

According to the covariance matching technology, the observed noise covariance can be updated in real time, the influence of the observed noise covariance on a filtering algorithm is reduced, and the new measurement value obtained by weight distribution is subjected to self-adaptive Kalman filtering. When facing time-varying information, the filtering algorithm can still achieve an ideal effect, so that the interference of noise covariance errors on positioning is overcome, and the positioning accuracy of the magnetic suspension train is improved.

In a possible implementation manner, referring to fig. 2, fig. 2 shows a flowchart of a method for determining a candidate position observation value of the target train in a train positioning system according to a first observability degree of the first positioning subsystem at each time and a second observability degree of the second positioning subsystem at each time according to a first embodiment of the present invention, which includes steps S201 to S203:

S201: and determining a first target position observation value of the first positioning subsystem at each time according to the first observability value of the first positioning subsystem at each time and the first initial position observation value of the target train at each time.

S202: and determining a second target position observation value of the second positioning subsystem at each time according to the second observability value of the second positioning subsystem at each time and the second initial position observation value of the target train at each time.

S203: and determining a candidate position observation value of the target train at each time according to the first target position observation value of the first positioning subsystem at each time and the second target position observation value of the second positioning subsystem at each time.

Specifically, reference is made to the specific algorithm in equation (11).

In a possible embodiment, the determining the first target position observation value of the first positioning subsystem at each time according to the first observability value of the first positioning subsystem at each time and the first initial position observation value of the target train at each time includes;

For each moment, determining the product of a first observable of the first positioning subsystem at the moment and a first initial position observed value of the target train at the moment as a first target position observed value of the first positioning subsystem at the moment.

for each time instant, determining the product of a second observability value of the second positioning subsystem at the time instant and a second initial position observation value of the target train at the time instant as a second target position observation value of the second positioning subsystem at the time instant.

In a possible implementation manner, after filtering candidate position observations of the target train at each time point in the train positioning system by using a unscented kalman filtering algorithm to obtain target position information of the target train at each time point, the method further comprises the steps of;

Specifically, for each moment, generating a coordinate point according to the target position information of the target train at the moment and the time of the moment, and performing curve fitting on each coordinate point to obtain the running track of the target train.

In addition to this, the above procedure is again described in supplementary detail here.

In order to overcome the defects of single sensor, easy interference, low positioning precision and the like of the traditional magnetic levitation train speed measurement and positioning method, the invention provides a self-adaptive unscented Kalman filtering method based on observability. Firstly, modeling an error state of a combined navigation system, calculating the observability of each subsystem according to a derived observability formula, and evaluating the information quality and performance of each subsystem; secondly, correcting the observed noise covariance matrix in real time by adopting a covariance matching method; and finally, information fusion is carried out by using the self-adaptive Kalman filtering algorithm. By comparing it with a standard unscented kalman filter algorithm, an improved algorithm positioning advantage is exhibited.

And integrating the acceleration and the rotation speed of the SINS sensor once, calculating the speed and the gesture, and integrating the speed to obtain the position. The navigation equations need to obtain initial values of position, velocity and attitude. This process is called initialization of position and velocity and calibration of pose.

1. Processing of combined system measurement data

For SINS, the purpose is of course very simple, deriving its navigation results (position, velocity, attitude) from the IMU raw angular velocity (gyroscope acquisition) and acceleration (accelerometer acquisition) data.

The inertial navigation system is a navigation parameter resolving system which uses a gyroscope and an accelerometer as sensitive devices, the system establishes a navigation coordinate system according to the output of the gyroscope, and the velocity and the position of a carrier in the navigation coordinate system are resolved according to the output of the accelerometer. The gyroscope in the inertial navigation system is used for forming a navigation coordinate system, so that the measuring axis of the accelerometer is stabilized in the coordinate system, and a course and an attitude angle are given; the accelerometer is used for measuring the acceleration of the moving body, the speed is obtained through one integration of time, and the distance is obtained through one integration of time.

The inertial navigation system realizes the basis of navigation positioning as follows:

wherein x (t) _k ) Represents the displacement at time k, x ₀ Represents initial displacement, t _k Represents the time at time k, t ₀ Represents the initial time, v (t) represents the velocity equation, v (t) _k ) Represents the velocity at time k, v (t ₀ ) The initial velocity is expressed, a (t) is the acceleration equation, and t is time.

The inertial navigation system at time t can be obtained according to equation (55) _k So that the position at time t can be solved according to equation (54).

In the invention, the SINS/GNSS chip is arranged on the train, and the longitude and latitude position information of the train is directly obtained by analyzing the received SINS/GNSS chip. The mileage of the maglev train is calculated by using longitude and latitude information. In the invention, the distance between adjacent latitude and longitude in unit time is calculated by utilizing the Haverin formula, and the total mileage is regarded as the mileage of the permanent magnetic levitation train.

/>

Wherein, the radius of the earth is represented,and->Represents the latitude, theta of two points A, B _A And theta _B Representing the longitude of A, B two points, d being the distance between A, B two points, hav representing the abbreviation of the semi-orthometric function.

Compared with SINS and GNSS, the Tag belongs to absolute positioning, so that the position data of the magnetic levitation train can be directly obtained, and the data do not need to be processed.

In the integrated navigation subsystem DOO calculation, if the DOO value is higher for one system, under the same condition, the higher the estimation accuracy is, the faster the convergence speed of the estimation error is. Each error state in the integrated navigation system has a unique observable measure and convergence speed. Thus, DOO value criteria may be applied to measure the observable ability of each error state.

The observability measure is for a certain state component at a certain moment, which means the ratio of the standard deviation of the initial setting error at moment 0 of the certain state component to the standard deviation of the filtering error of the same state component at moment k. The larger the number of observables is, the more remarkable the estimation error reduction degree of the corresponding state components is or the more remarkable the precision improving effect is after Kalman filtering.

Human set observability D of approximate judgment State component _k(j) The strength is as follows:

D _k(j) not more than 1 is not observable

1< _k(j) Less than or equal to 2 weakness

2< _k(j) Middle-grade less than or equal to 10

D _k(j) >10 strong (58)

The calculated value of DOO is a scalar. This feature differs from other existing forms in that it allows the DOO of each state variable to be directly calculated without eigenvalues and singular values. In addition, DOO can be easily calculated by using a Kalman filtering algorithm, and the use mode with smaller calculation amount is simpler than other methods.

To calculate the observability of the integrated navigation system, the SINS error model is taken as the system model, and the positioning and velocity errors are measured values. The integrated navigation system consists of two independent parts, an SINS/GNSS subsystem and an SINS/Tag subsystem. In most cases, each sensor operates independently, and its operating conditions may change due to environmental changes. To evaluate the performance of these three subsystems, DOO is used as an index, and a scalar form represented by (10) is used for calculation, so that the information quality and performance index of each subsystem are evaluated according to the calculated DOO value.

Judging the observability intensity of the state component according to the approximate setting by people, when the DOO value of the calculated state component is less than or equal to 1, indicating that the state component is not observable, and directly removing the state component without participating in filtering in order to avoid reducing the positioning precision and diverging the unscented Kalman filtering; when the observability of the state component is weak or medium, the duty ratio of the current state component in the filtering algorithm can be reduced, the influence of the state component on the filtering positioning precision is reduced, and the probability of filtering divergence is reduced; if the observability of the currently calculated state component is strong, the positioning error of the state component is proved to be small, the positioning accuracy is higher, and the state component can be used as the main part in the filtering algorithm in the whole filtering process, so that the filtering error is reduced, and the positioning accuracy and fault tolerance of the permanent magnetic levitation train are improved.

And (3) obtaining the DOO value of the subsystem according to the step (10), distributing corresponding weights to the calculated DOO value of the subsystem, and substituting the calculated DOO value of the subsystem into the step (11) to obtain a new observed value, so that the obtained new observed value is filtered.

In adaptive UKF information fusion, the performance of standard UKF unscented Kalman filtering depends on the exact statistical characteristics of the system noise. If the noise distribution of the inertial navigation system and the global navigation satellite system receiver is not corrected in real time, the standard UKF filtering effect will greatly reduce even the divergence state. The invention provides a self-adaptive UKF for correcting observed noise covariance in real time so as to overcome the limitation of a standard UKF. According to covariance matching technology, a covariance matrix of observed noise is determined by using the innovation sequence. The proposed algorithm can estimate and adjust the system noise statistics online, thereby enhancing the adaptive capacity of the standard UKF.

And (3) carrying out self-adaptive UKF algorithm fusion on the new observed value obtained in the step (11) and corresponding data, wherein in the fusion process, if the observed noise covariance matrix is unmatched with the current filtering, the current observed data is more important, the current observed noise covariance matrix needs to be updated to adapt to the current filtering process, and the optimal forgetting factor is introduced in the process of recalculating the noise covariance matrix, and noise statistical data is estimated and updated on line, so that the error of time-varying noise on system state estimation is reduced, and the positioning precision of the permanent magnetic levitation train system is improved.

Compared with a standard UKF and a strong tracking UKF, the proposed self-adaptive UKF calculates the DOO value of each combined subsystem based on observability, evaluates the information quality and the working performance of the combined subsystem, and distributes weight for the observed value to prevent the situation of error increase and even divergence in the filtering process. And the average error (ME), the average relative error (MRE) and the Root Mean Square Error (RMSE) of the self-adaptive UKF are reduced, compared with the real track of the magnetic levitation train, the track predicted by the self-adaptive UKF is closer, the error is smaller, and the invention adopts a filtering method to reduce the positioning error and improve the positioning precision of the permanent magnetic levitation train.

Example two

Referring to fig. 3, fig. 3 shows a schematic structural diagram of a train positioning device based on a kalman filtering algorithm according to a second embodiment of the present invention, where the device includes:

the position information determining module 301 is configured to obtain first position information of the target train in the SINS inertial navigation system at each time, second position information of the target train in the GNSS global navigation satellite system, and third position information of the target train acquired by Tag tags disposed on the target train;

A first initial position observation value determining module 302, configured to determine, according to the first position information and the second position information at each time, a first initial position observation value of the target train at each time in a first positioning subsystem by using a kalman filtering algorithm, where the first positioning subsystem is composed of the inertial navigation system and the global navigation satellite system;

the second initial position observation value determining module 303 determines a second initial position observation value of the target train at each time point in a second positioning subsystem according to the first position information and the third position information at each time point by using a kalman filtering algorithm, wherein the second positioning subsystem consists of the inertial navigation system and the Tag;

an observability determining module 304, configured to determine a first observability degree of the first positioning subsystem at each time point according to a state vector of a first initial position observation value of the target train at each time point and observation noise of the first initial position observation value of the target train at each time point, and determine a second observability degree of the second positioning subsystem at each time point according to a state vector of a second initial position observation value of the target train at each time point and observation noise of the second initial position observation value of the target train at each time point;

A candidate position observation value determining module 305, configured to determine a candidate position observation value of the target train in a train positioning system at each time according to a first observability degree of the first positioning subsystem at each time and a second observability degree of the second positioning subsystem at each time, where the train positioning system is composed of the inertial navigation system, the global navigation satellite system and the Tag;

the target position information determining module 306 is configured to utilize an unscented kalman filtering algorithm to filter candidate position observations of the target train at each time in the train positioning system to obtain target position information of the target train at each time.

In a possible embodiment, the candidate position observation determining module is specifically configured to, when determining the candidate position observation of the target train in the train positioning system at each time point according to the first observability degree of the first positioning subsystem at each time point and the second observability degree of the second positioning subsystem at each time point:

In a possible embodiment, the candidate position observation determining module is specifically configured to, when determining the first target position observation value of the first positioning subsystem at each time point according to the first observability degree of the first positioning subsystem at each time point and the first initial position observation value of the target train at each time point:

In a possible implementation manner, referring to fig. 4, fig. 4 shows a schematic structural diagram of a second train positioning device based on a kalman filtering algorithm according to a second embodiment of the present application, where the device further includes:

and the running track determining module 401 is configured to, after the target position information determining module filters candidate position observations of the target train at each time point in the train positioning system by using a unscented kalman filtering algorithm to obtain target position information of the target train at each time point, generate a running track of the target train according to the target position information of the target train at each time point and time information of each time point.

Example III

Based on the same application concept, referring to fig. 5, fig. 5 shows a schematic structural diagram of a computer device provided by a third embodiment of the present application, where, as shown in fig. 5, a computer device 500 provided by the third embodiment of the present application includes:

the system comprises a processor 501, a memory 502 and a bus 503, wherein the memory 502 stores machine-readable instructions executable by the processor 501, when the computer device 500 is running, the processor 501 and the memory 502 communicate through the bus 503, and the machine-readable instructions are executed by the processor 501 to perform the steps of the train positioning method based on the kalman filter algorithm as shown in the second embodiment.

Example IV

Based on the same application concept, the embodiment of the present invention further provides a computer readable storage medium, where a computer program is stored, and when the computer program is executed by a processor, the steps of the train positioning method based on the kalman filter algorithm in any one of the above embodiments are executed.

It will be clear to those skilled in the art that, for convenience and brevity of description, specific working procedures of the above-described system and apparatus may refer to corresponding procedures in the foregoing method embodiments, which are not described herein again.

The computer program product for performing the train positioning based on the kalman filter algorithm provided by the embodiment of the invention comprises a computer readable storage medium storing program codes, wherein the instructions included in the program codes can be used for executing the method described in the method embodiment, and specific implementation can be referred to the method embodiment and will not be repeated here.

The train positioning device based on the Kalman filtering algorithm provided by the embodiment of the invention can be specific hardware on equipment or software or firmware installed on the equipment. The device provided by the embodiment of the present invention has the same implementation principle and technical effects as those of the foregoing method embodiment, and for the sake of brevity, reference may be made to the corresponding content in the foregoing method embodiment where the device embodiment is not mentioned. It will be clear to those skilled in the art that, for convenience and brevity, the specific operation of the system, apparatus and unit described above may refer to the corresponding process in the above method embodiment, which is not described in detail herein.

In the embodiments provided in the present invention, it should be understood that the disclosed apparatus and method may be implemented in other manners. The above-described apparatus embodiments are merely illustrative, for example, the division of the units is merely a logical function division, and there may be other manners of division in actual implementation, and for example, multiple units or components may be combined or integrated into another system, or some features may be omitted, or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed with each other may be through some communication interface, device or unit indirect coupling or communication connection, which may be in electrical, mechanical or other form.

The units described as separate units may or may not be physically separate, and units shown as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

In addition, each functional unit in the embodiments provided in the present invention may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit.

It should be noted that: like reference numerals and letters in the following figures denote like items, and thus once an item is defined in one figure, no further definition or explanation of it is required in the following figures, and furthermore, the terms "first," "second," "third," etc. are used merely to distinguish one description from another and are not to be construed as indicating or implying relative importance.

Finally, it should be noted that: the above examples are only specific embodiments of the present invention, and are not intended to limit the scope of the present invention, but it should be understood by those skilled in the art that the present invention is not limited thereto, and that the present invention is described in detail with reference to the foregoing examples: any person skilled in the art may modify or easily conceive of the technical solution described in the foregoing embodiments, or perform equivalent substitution of some of the technical features, while remaining within the technical scope of the present disclosure; such modifications, changes or substitutions do not depart from the spirit and scope of the corresponding technical solutions. Are intended to be encompassed within the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.

Claims

1. A method for locating a train based on a kalman filter algorithm, the method comprising:

respectively acquiring first position information of a target train in an SINS inertial navigation system at each moment, second position information of the target train in a GNSS global navigation satellite system and third position information of the target train acquired by Tag labels arranged on the target train;

2. The method of claim 1, wherein the determining the candidate location observations of the target train at each time instant in the train positioning system based on the first observability of the first positioning subsystem at each time instant and the second observability of the second positioning subsystem at each time instant comprises:

3. The method of claim 2, wherein the determining the first target position observation for the first positioning subsystem at each time instant from the first observable for the first positioning subsystem at each time instant and the first initial position observation for the target train at each time instant comprises;

4. The method of claim 1, wherein after filtering candidate position observations of the target train at each time point in the train positioning system by using an unscented kalman filter algorithm to obtain target position information of the target train at each time point, the method further comprises;

5. A kalman filter algorithm-based train positioning device, the device comprising:

the position information determining module is used for respectively acquiring first position information of a target train in the SINS inertial navigation system at each moment, second position information of the target train in the GNSS global navigation satellite system and third position information of the target train acquired by Tag labels arranged on the target train;

6. The apparatus of claim 5, wherein the candidate position fix module, when configured to determine the candidate position fix for the target train in the train positioning system at each time instant based on the first observability of the first positioning subsystem at each time instant and the second observability of the second positioning subsystem at each time instant, is specifically configured to:

7. The apparatus of claim 6, wherein the candidate position fix module, when configured to determine the first target position fix for the first positioning subsystem at each time instant based on the first observable for the first positioning subsystem at each time instant and the first initial position fix for the target train at each time instant, is specifically configured to:

8. The apparatus of claim 5, wherein the apparatus further comprises:

9. A computer device, comprising: a processor, a memory and a bus, said memory storing machine readable instructions executable by said processor, said processor and said memory communicating via the bus when the computer device is running, said machine readable instructions when executed by said processor performing the steps of the kalman filter algorithm based train positioning method according to any of claims 1 to 4.

10. A computer readable storage medium, characterized in that the computer readable storage medium has stored thereon a computer program which, when run by a processor, performs the steps of the kalman filter algorithm based train positioning method according to any of claims 1 to 4.