Railway map matching method and system based on Beidou navigation system and odometer

Abstract

The invention provides a railway map matching method and system based on a Beidou navigation system and an odometer, and relates to the technical field of heavy haul railway transportation, and the method comprises the steps: obtaining first state data of a train based on the Beidou navigation system, and obtaining second state data of the train based on the odometer; carrying out fusion processing on the first state data and the second state data, and determining observation states of the train at all moments and corresponding state transition probabilities and state emission probabilities; and determining an optimal matching path of the train at each moment based on the observation state of the train at each moment and the corresponding state transition probability and state emission probability. According to the method and the system provided by the invention, the positioning precision and the stability of the heavy haul railway vehicle in a complex environment are improved.

Description

Technical Field

[0001] The present application relates to the field of heavy-haul railway transportation technology, and in particular to a railway map matching method and system based on a Beidou navigation system and an odometer. Background Art

[0002] With the increasing demand for positioning accuracy in heavy-haul railway systems and track determination technology, map matching technology has been widely studied.

[0003] Railway map matching methods in related technologies mainly focus on tightly coupled positioning methods using the Global Navigation Satellite System (GNSS) and the Inertial Navigation System (INS). When encountering complex heavy-load railway environments, such as tunnels or densely built-up areas in cities, GNSS will generate high noise, and positioning accuracy and stability will be greatly reduced. In addition, these methods mainly focus on road-level matching rather than section-level matching. In situations where precise lane control and train safety management are required, road-level matching may lead to safety risks.

[0004] Therefore, how to improve the positioning accuracy and stability of heavy-load railway vehicles in complex environments has become a technical problem that needs to be urgently solved in the industry. Summary of the invention

[0005] The present application provides a railway map matching method and system based on the Beidou navigation system and the odometer, which are used to solve the technical problem of how to improve the positioning accuracy and stability of heavy-load railway vehicles in complex environments.

[0006] The present application provides a railway map matching method based on the Beidou navigation system and an odometer, comprising: Acquiring first status data of a train based on a Beidou navigation system, and acquiring second status data of the train based on an odometer; Performing fusion processing on the first state data and the second state data to determine the observed state of the train at each time and the corresponding state transition probability and state emission probability; Based on the observed state of the train at each moment and the corresponding state transition probability and state emission probability, the optimal matching path of the train at each moment is determined.

[0007] In some embodiments, the fusing of the first state data and the second state data to determine the observed state of the train at each time and the corresponding state transition probability and state emission probability includes: Performing fusion processing on the first state data and the second state data based on a Kalman filter to obtain state fusion data; The state fusion data is processed based on a hidden Markov model and a Viterbi algorithm to obtain the observed state of the train at each moment and the corresponding state transition probability and state emission probability.

[0008] In some embodiments, before fusing the first state data and the second state data based on the Kalman filter to obtain state fusion data, the method further includes: The first state data and the second state data are preprocessed; the preprocessing includes at least one of data standardization, time synchronization and filtering processing.

[0009] In some embodiments, the state fusion data is processed based on a hidden Markov model and a Viterbi algorithm to obtain the observed state of the train at each time and the corresponding state transition probability and state emission probability, including: Processing the state fusion data based on a hidden Markov model, establishing a relationship between the state fusion data and the possible state of the train, and determining the possible state of the train at each moment and the corresponding state transition probability and state emission probability; Based on the Viterbi algorithm, possible states of the train at each moment and corresponding state transition probabilities and state emission probabilities are recursively calculated to determine the observed state of the train at each moment.

[0010] In some embodiments, the method further comprises: Dividing the travel route of the train into a plurality of initial segments, and determining the segmentation state of each initial segment; the segmentation state includes an observed state of the train passing through the initial segment; Based on the segment states of the initial segments and the corresponding state transition probabilities and state emission probabilities, the optimal path segment of the travel road of the train is determined.

[0011] In some embodiments, determining the optimal path segment of the travel road of the train based on the segment states of each initial segment and the corresponding state transition probability and state emission probability includes: Based on the Viterbi algorithm, the segment states of each initial segment and the corresponding state transition probability and state emission probability are recursively calculated to obtain the optimal path segment of the travel road of the train.

[0012] In some embodiments, the method further comprises: Analyzing the trajectory data of the train, and determining the angle between adjacent line segments based on the angles of adjacent line segments; Determining the distance between the adjacent line segments based on the coordinates of the starting points or the end points of the adjacent line segments; Determining a degree of trajectory splitting of the adjacent line segments based on the slopes of the adjacent line segments; Determine a penalty factor for merging intersection points based on the angle and distance between the adjacent line segments; Determining a non-merged intersection penalty factor based on the angle, distance and trajectory splitting degree of the adjacent line segments; Based on the merged intersection penalty factor and the non-merged intersection penalty factor, the selection probability of the intersection of the adjacent line segments is determined.

[0013] The present application provides a railway map matching system based on the Beidou navigation system and an odometer, comprising: A data acquisition module, used for acquiring first status data of a train based on a Beidou navigation system, and acquiring second status data of the train based on an odometer; A data fusion module, used for fusing the first state data and the second state data to determine the observed state of the train at each time and the corresponding state transition probability and state emission probability; The map matching module is used to determine the optimal matching path of the train at each moment based on the observed state of the train at each moment and the corresponding state transition probability and state emission probability.

[0014] The present application provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the railway map matching method based on the Beidou navigation system and the odometer is implemented.

[0015] The present application provides a non-transitory computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the railway map matching method based on the Beidou navigation system and the odometer is implemented.

[0016] The railway map matching method and system based on Beidou navigation system and odometer provided in the present application obtains the first state data of the train based on the Beidou navigation system, and obtains the second state data of the train based on the odometer; the first state data and the second state data are fused and processed to determine the observation state of the train at each time and the corresponding state transition probability and state emission probability; based on the observation state of the train at each time and the corresponding state transition probability and state emission probability, the optimal matching path of the train at each time is determined; by efficiently integrating and analyzing the multi-source data of Beidou navigation system and odometer, it is ensured that these data are fully utilized and optimized in the map matching process; according to the observation state of the train at each time and the corresponding state transition probability and state emission probability, the optimal matching path of the train at each time is determined, and the adaptability of map matching in different heavy-duty railway network structures and changing environmental conditions is enhanced through improved algorithm design, ensuring that high-precision and high-reliability positioning services can be provided even when the Beidou signal is weak or invalid; the positioning accuracy and stability of heavy-duty railway vehicles in complex environments are improved. In addition, the problem that the current mainstream map matching algorithm cannot be accurate to the section level is solved, and the segmentation logic of heavy-duty railway lines is optimized. The improved hidden Markov model algorithm is used to segment heavy-haul railway lines more accurately and intelligently match path segments, which significantly improves the matching accuracy and overall performance of the heavy-haul railway system in complex scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0018] In order to more clearly illustrate the technical solutions in the present application or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0019] Figure 1 It is a flow chart of the railway map matching method based on the Beidou navigation system and odometer provided in this application.

[0020] Figure 2 It is a structural schematic diagram of a railway map matching system based on the Beidou navigation system and odometer provided in this application.

[0021] Figure 3 It is a schematic diagram of the overall solution of the railway map matching method based on the Beidou navigation system and odometer provided in this application.

[0022] Figure 4It is a data fusion schematic diagram of the railway map matching method based on the Beidou navigation system and odometer provided in this application.

[0023] Figure 5 It is a structural schematic diagram of the electronic device provided by this application. DETAILED DESCRIPTION

[0024] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present application.

[0025] It should be noted that the terms "first", "second" etc. in the present application are used to distinguish similar objects, and need not be used to describe a specific order or sequential order. It should be understood that the data used in this way can be interchangeable in appropriate circumstances, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, the process, method, system, product or equipment comprising a series of steps or units or modules need not be limited to those steps or units or modules clearly listed, but may include other steps or units or modules that are not clearly listed or inherent to these processes, methods, products or equipment.

[0026] The importance of map matching technology is particularly prominent in complex geographical environments, such as mountainous areas or tunnels, where GNSS signals may be extremely unstable or completely lost. Therefore, map matching technology must be able to adapt to this harsh operating environment to ensure accurate location identification and trajectory tracking even under extreme conditions. At the same time, the frequent loading and unloading activities of heavy-duty railways and extreme operating conditions such as high-frequency vibrations and temperature changes pose special challenges to map matching technology. Mechanical vibrations during loading and unloading may affect the signal quality of GNSS and navigation sensors, and long-term exposure to harsh environments will cause the performance of the Inertial Measurement Unit (IMU) to gradually degrade, manifested as sensor drift and increased noise. Therefore, map matching algorithms need to be highly robust and adaptable.

[0027] The high-precision positioning of heavy-duty railways in related technologies relies on a single or a few positioning technologies, especially GNSS or systems tightly coupled with INS, which can provide high positioning accuracy in areas with a wide field of view, but their performance drops sharply in environments with obstructed vision, such as tunnels or areas with tall buildings in cities. In addition, although the tightly coupled system can theoretically enhance the accuracy and robustness of positioning, in actual applications it may be negatively affected by the instability of GNSS signal quality and the accumulation of INS errors.

[0028] The map matching algorithm in the related art mainly relies on GNSS data, while ignoring other potential auxiliary information sources, such as odometer and gyroscope data. This single data source usage strategy increases the risk of map matching errors when the GNSS signal is unstable or lost.

[0029] The map matching algorithms in related technologies lack flexibility in adapting to changing environmental conditions and complex heavy-duty railway network structures, especially when dealing with special situations such as U-turn mismatches, and lack effective solutions. At the same time, in heavy-duty railway systems, current high-precision positioning technologies have not been widely used to achieve accurate matching at the section level, which has caused serious hidden dangers in the safety and dispatching efficiency of heavy-duty railways. Especially in complex heavy-duty railway networks and high-speed operation conditions, the inability to accurately identify the position of trains on specific sections may lead to dispatching errors, increase the risk of train collisions and derailments, and thus pose a major threat to passenger safety.

[0030] In order to solve the shortcomings of related technologies, Figure 1 This is a flow chart of a railway map matching method based on the Beidou navigation system and odometer provided by this application, such as Figure 1 As shown, the method includes step 110 , step 120 and step 130 .

[0031] Step 110: Acquire first status data of the train based on the Beidou navigation system, and acquire second status data of the train based on the odometer.

[0032] Specifically, the method provided in the embodiment of the present application is implemented by a railway map matching system. The system can be implemented by software, such as a railway map matching program running on a computer or server; or by hardware, such as a mobile terminal, computer or server that executes the railway map matching method.

[0033] The method provided in the embodiment of the present application sets a Beidou navigation system positioning sensor and an odometer on the train. The positioning sensor receives the signal sent by the Beidou navigation satellite and generates first state data, which can represent the position, speed, attitude and other information of the train. The odometer usually records the cumulative distance traveled by the train from a certain reference point, and can generate second state data, which can represent the position and speed and other information of the train.

[0034] Step 120: The first state data and the second state data are fused to determine the observed state of the train at each moment and the corresponding state transition probability and state emission probability.

[0035] Specifically, the first state data and the second state data each have advantages. The first state data and the second state data can be fused. Using Beidou signals in open environments and using odometer data in shielded environments such as cities and tunnels effectively overcomes the positioning problem of the Beidou navigation system in high-noise environments, ensures the accuracy and reliability of the map matching system in a variety of environments such as cities and tunnels, and thus achieves accurate positioning of heavy-load railway vehicles.

[0036] The fusion processing method can be combined with the hidden Markov model and the Viterbi algorithm. The hidden Markov model (HMM) is a statistical model used to describe a random process with hidden states, which can be used to infer the most likely state sequence for a given observation sequence. The Viterbi algorithm is a dynamic programming algorithm used to find the most likely state sequence for a given observation sequence in the hidden Markov model.

[0037] The observed state refers to events or data that can be directly observed or measured. In the embodiment of the present application, the observed state of the train at each moment refers to information such as the position, speed, and acceleration of the train at each moment.

[0038] The state transition probability refers to the probability that the train transfers from one state to another. The state emission probability is an important parameter that describes the relationship between the hidden state and the observed state. It indicates the probability of generating a specific observed state under a specific hidden state.

[0039] Step 130: Determine the optimal matching path of the train at each time based on the observed state of the train at each time and the corresponding state transition probability and state emission probability.

[0040] Specifically, the Newson algorithm is a method for analyzing time series data. The Newson algorithm can be improved to perform map matching in the embodiments of the present application.

[0041] Combining the Newson algorithm with the hidden Markov model can take advantage of the strengths of both to process complex time series data. Use the Newson algorithm to decompose the time series, extract trends and random components, and then use the hidden Markov model to model these components. The hidden Markov model can capture potential hidden state changes, while the Newson algorithm helps identify and model obvious patterns. The Newson algorithm can provide rich information about the time series, which can be used as observations for the HMM, thereby enhancing the HMM's ability to model hidden states.

[0042] By dynamically adjusting the weights of transition probability and observation probability, the improved Newson algorithm can effectively handle the uncertainty of segment selection and reduce the possibility of false matching, thereby improving the accuracy and robustness of matching. These features make the improved Newson algorithm a powerful tool for complex traffic system analysis and applications that require high-precision map matching.

[0043] The railway map matching method based on the Beidou navigation system and the odometer provided in the embodiment of the present application obtains the first state data of the train based on the Beidou navigation system, and obtains the second state data of the train based on the odometer; the first state data and the second state data are fused and processed to determine the observed state of the train at each time and the corresponding state transition probability and state emission probability; based on the observed state of the train at each time and the corresponding state transition probability and state emission probability, the optimal matching path of the train at each time is determined; by efficiently fusing and analyzing the multi-source data of the Beidou navigation system and the odometer, it is ensured that these data are fully utilized and optimized in the map matching process; according to the observed state of the train at each time and the corresponding state transition probability and state emission probability, the optimal matching path of the train at each time is determined, and the adaptability of map matching in different heavy-load railway network structures and changing environmental conditions is enhanced through improved algorithm design, ensuring that high-precision and high-reliability positioning services can be provided even when the Beidou signal is weak or fails; the positioning accuracy and stability of heavy-load railway vehicles in complex environments are improved.

[0044] It should be noted that each implementation method of the present application can be freely combined, the order can be changed, or it can be executed separately, and does not need to rely on or depend on a fixed execution order.

[0045] In some embodiments, the first state data and the second state data are fused to determine the observed state of the train at each time and the corresponding state transition probability and state emission probability, including: Performing fusion processing on the first state data and the second state data based on the Kalman filter to obtain state fusion data; The state fusion data is processed based on the hidden Markov model and Viterbi algorithm to obtain the observed state of the train at each moment and the corresponding state transition probability and state emission probability.

[0046] Specifically, data is collected from two sensors, the Beidou navigation system sensor and the odometer, which represent the position, speed and attitude information of the vehicle (first state data and second state data). The collected data is first fused through a Kalman filter, which fuses the first state data and the second state data to achieve a more comprehensive state estimation.

[0047] After that, the fused data is processed using hidden Markov models and Viterbi algorithms to infer the most likely sequence of vehicle states, thereby more accurately predicting the vehicle's future position and driving state. For these data, Beidou data is usually used in open areas, and odometer data is used to correct errors, while odometer data is mainly used in high-occlusion environments such as urban environments and mountain tunnels. This series of operations effectively utilizes multi-source sensor data to achieve high-precision positioning and monitoring.

[0048] The railway map matching method based on the Beidou navigation system and the odometer provided in the embodiment of the present application utilizes a Kalman filter to fuse and optimize the data of various sensors, thereby ensuring the accuracy and consistency of the data; the hidden Markov model and the Viterbi algorithm are used to refine the data and optimize the map matching process, thereby significantly improving the estimation accuracy of positioning and motion status.

[0049] In some embodiments, before fusing the first state data and the second state data based on the Kalman filter to obtain the state fusion data, the method further includes: The first state data and the second state data are preprocessed; the preprocessing includes at least one of data standardization, time synchronization and filtering processing.

[0050] Specifically, data standardization refers to converting data with different characteristics to the same scale to avoid the impact of dimensional differences on the model. Time synchronization refers to ensuring that time series data from different sources can be compared and analyzed at the same time point. Filtering refers to removing noise and unnecessary fluctuations to extract the main trends and patterns of the signal.

[0051] The railway map matching method based on the Beidou navigation system and the odometer provided in the embodiment of the present application pre-processes the first state data and the second state data to ensure the accuracy and consistency of the data.

[0052] In some embodiments, the state fusion data is processed based on the hidden Markov model and the Viterbi algorithm to obtain the observed state of the train at each time and the corresponding state transition probability and state emission probability, including: The state fusion data is processed based on the hidden Markov model, the relationship between the state fusion data and the possible state of the train is established, and the possible state of the train at each moment and the corresponding state transition probability and state emission probability are determined; Based on the Viterbi algorithm, the possible states of the train at each moment and the corresponding state transition probabilities and state emission probabilities are recursively calculated to determine the observed state of the train at each moment.

[0053] Specifically, the calculation process of the above method is as follows: Step 1: Initialization

[0054] In the formula, Represents the initial GNSS position data, indicating the initial geographic location of the train, including longitude, latitude and altitude coordinates, as well as acquisition time and accuracy information.

[0055] The initial GNSS position data is the initial position data obtained through the Global Navigation Satellite System (GNSS), including longitude, latitude and altitude. These data are used to determine the precise position of the train on the surface of the earth. Specifically, it includes longitude (the east / west coordinates of the train's location, usually in degrees), latitude (the south / north coordinates of the train's location, usually in degrees), altitude (the height of the train relative to sea level, usually in meters), timestamp (the specific time when the GNSS data is obtained, usually in UTC time format) and accuracy information (including horizontal accuracy and vertical accuracy, indicating the accuracy of the positioning data).

[0056] Represents the initial odometer data, indicating the cumulative distance and speed traveled by the train from the reference point, and the acquisition time.

[0057] Initial odometer data is the initial distance data obtained from the train's odometer (or odometer). Odometer data usually records the cumulative distance traveled by the train from a reference point. Specifically, it includes cumulative distance traveled (the total distance traveled by the train from the reference point, usually in meters or kilometers), speed (the current speed of the train, usually in meters per second or kilometers per hour) and timestamp (the specific time when the odometer data is obtained, usually in UTC format).

[0058] represents the state probability distribution based on the initial data, which represents the probability distribution of various states that the train may be in at the initial moment, including the initial probability values ​​of position, speed and acceleration states; represents probability; Indicates the initial state; Indicates status.

[0059] The state probability distribution based on the initial data is a probability distribution, which indicates the probability that the train may be in each state at the initial moment. The state usually includes the specific position, speed, acceleration and other information of the train. Specifically, it includes a state set and an initial probability value. The state set refers to the set of all states that the train may be in, and each state includes the position state (longitude, latitude and altitude coordinates), speed state (speed of the train) and acceleration state (acceleration of the train). The initial probability value refers to the initial probability of each state, indicating the possibility of the train being in that state.

[0060] Step 2: Calculate state transition

[0061] The input of this step is the last moment The state probability distribution , current moment Odometer reading and the current moment Beidou location data The output is the current time The state probability distribution . It can be expressed by the formula:

[0062] In the formula, Indicates time state The probability distribution of , is a state variable, indicating the position, speed, acceleration, etc. of the train; For the moment odometer reading; For the moment Beidou data; represents the state transition probability.

[0063]

[0064] In the formula, is the normalization constant; is the smoothing parameter; is the state of the previous moment, indicating that the train is at time Position, speed and acceleration (specifically the longitude, latitude, altitude, speed and acceleration of the previous moment); is the current state, indicating that the train is at time The position, speed and acceleration of the object (specifically the longitude, latitude, altitude, speed and acceleration at the current moment).

[0065] is a fusion function that combines the deviation between the prediction and the actual change of the odometer increment and the Beidou position, specifically:

[0066] In the formula, for The predicted position of the axis, for The actual position of the axis; for The predicted position of the axis, for The actual position of the axis; for The predicted position of the axis, for The actual position of the axis.

[0067] Step 3: Calculate the emission probability

[0068] The input of this step is the current time Beidou location data , current moment Odometer reading and the current moment Status The output is the emission probability . It can be expressed by the formula:

[0069] In the formula, the emission probability Indicates the current state Under, observe the current Beidou and odometer data It reflects the consistency between the observed data and the state. It is the current observation data, including the current Beidou data and odometer data, which indicates the train position and travel distance data observed by the sensor at the current moment. is the smoothing parameter.

[0070] Calculate the distance between the current Beidou position and the trajectory position predicted by the state St, which can be expressed as:

[0071] In the formula, For Beidou data Axis position, Status Predicted Axis trajectory position; For Beidou data Axis position, Status Predicted Axis trajectory position; For Beidou data Axis position, Status Predicted Axis trajectory position.

[0072] Step 4: Update and predict

[0073] The input of this step is the transition probability of all current states , emission probability The output is the updated state probability.

[0074] In this step, the Viterbi Algorithm is applied to update the state probability according to the above transfer and emission probabilities to find the most likely state sequence. The state sequence represents the order of all states experienced by the train over a period of time. It includes the specific position, speed, acceleration and other information of the train at each moment, through which the movement trajectory and state changes of the train over the entire time period can be tracked. It can be expressed as:

[0075] In the formula, Indicates at time Moment, state The probability of the most likely state sequence at the end. This value is obtained by maximizing the probability of the most likely state sequence at the previous moment, the state transition probability, and the emission probability. In other words, at the current moment The train is in state The probability of the optimal path.

[0076] In the recursive calculation process of the Viterbi algorithm, Indicates at time The train's status It is used to calculate from time To time The state transition probability, in the formula, j represents all possible states at the previous moment.

[0077] Indicates at the current moment Data observed by sensors (such as Beidou, odometer, etc.). It is the actual observation value at the current moment, usually acquired by multiple sensors and combined to form a complete observation data set.

[0078] For the current moment It includes information such as train position, travel distance, speed, etc. observed by Beidou and odometer sensors. These data are used to calculate the launch probability, that is, in the current state Observed below possibility.

[0079] The railway map matching method based on the Beidou navigation system and the odometer provided in the embodiment of the present application processes the state fusion data according to the hidden Markov model and the Viterbi algorithm to obtain the observed state of the train at each moment and the corresponding state transition probability and state emission probability, which significantly improves the estimation accuracy of positioning and motion state.

[0080] In some embodiments, the advantage of the improved Newson algorithm is its high adaptability to complex road networks and vehicle behaviors. The algorithm not only considers the geometric structure and topological relationship of the road, but also integrates the time constraints and the motion characteristics of the vehicle, so that it can accurately match the vehicle position and path in a changing traffic environment.

[0081] The map matching calculation process of the improved Newson algorithm can be described by the following steps and mathematical expressions: Step 1: Initialization

[0082] The input to this step is the GNSS data of the initial position , initial odometer reading , the output is the probability distribution of the initial state The processing is to use the Beidou and odometer data of the initial position to determine the probability distribution of the initial state.

[0083] Step 2: Calculate state transition

[0084] The input of this step is the time The state probability distribution , the current odometer reading , the output is the time The state probability distribution , which can be expressed as:

[0085] In the formula, Indicates status and status Specifically, it is a function that calculates the straight-line distance between two state positions. It reflects the distance traveled by the train between the two states. This distance is used to calculate the state transition probability, which helps determine the likelihood of a train moving from one state (position) to another. This distance function is used to measure the extent of a train's movement between two states. By combining the odometer reading and the actual distance, the train's travel path and state changes over a given period of time can be estimated. It can be expressed as:

[0086] In the formula, Indicates status of Axis position coordinates, Indicates status of Axis position coordinates; Indicates status of Axis position coordinates, Indicates status The y-axis position coordinate of Indicates status The z-axis position coordinate, Indicates status The z-axis position coordinate of .

[0087] Step 3: Calculate the emission probability

[0088] The input of this step is the current time BeiDou data , current moment Status , the output is the emission probability , which can be expressed as:

[0089] Indicates status The position projected onto the road represents the expected position of the train on the road under a given state. Through this function, the state information can be converted into a position in the road coordinate system, which is used to calculate the distance from the actual observed position. This function is used to convert the current state into a position on the road, which is convenient for comparison with Beidou observation data to calculate the launch probability. The launch probability reflects the consistency between the current observation data and the state, which is crucial for accurate trajectory tracking and position prediction. It can be expressed as:

[0090] In the formula, Indicates status The x-axis, y-axis, and z-axis coordinates projected onto the road.

[0091] Step 4: Update and predict

[0092] The input of this step is the transition probability of all current states and the emission probability , the output is the updated state probability .

[0093] The process is to apply the Viterbi algorithm to update the state probabilities. The Viterbi algorithm is a dynamic programming algorithm used to find the most likely state sequence under a given observation sequence.

[0094] Input is the state transition probability matrix , the emission probability matrix , the observation sequence , initial state probability .

[0095] State transition probability matrix represents the probability of transitioning from one state to another. Specifically, the elements in the matrix Indicates at time Always in the state Then, transfer to time Status at all times probability.

[0096] Emission probability matrix represents the probability of observing a specific observation value in a given state. Specifically, the elements in the matrix Indicates in status Observed value In heavy-haul railway systems, the emission probability matrix is ​​used to describe the Beidou data and odometer data observed in different states. For example, the probability of observing specific Beidou position data in a certain position state.

[0097] Observation sequence represents the sequence of values ​​observed at different time points. Specifically, the observation sequence It contains a series of observation values, which are obtained at different time points through sensors such as Beidou and odometer.

[0098] The output is the most likely state sequence .

[0099] The above process can be expressed as:

[0100] represents the probability of the maximum probability path of the train in state i at the initial time (time t=1). This step initializes the probability of each state, which is calculated based on the initial state probability and the emission probability, and represents the probability value of the most likely path of the train in different states at the initial time.

[0101] represents the initial state probability, which represents the probability that the train is in state i at the initial moment, and represents the probability that the train occupies a specific state at the initial moment.

[0102] Indicates that the first observation value is observed in state i The probability of , represents the probability of observing the initial observation data in state i.

[0103] It is the path pointer, which is used to record the most likely previous state at the initial moment. In the initialization phase, this value is 0, because the initial state has no previous state, which represents the pointer for recording the path information, and is initialized to 0.

[0104] The recursive calculation process can be expressed as:

[0105] represents the probability of the maximum probability path of the train in state i at time t, and represents the probability value of the most likely path of the train in state i at time t.

[0106] Indicates from time The maximum probability value of each state j transferring to state i at time t represents the maximum path probability of finding the transfer from the previous moment to the current moment to a specific state.

[0107] is the state transition probability, which indicates the probability of transitioning from state j to state i, and represents the probability value of the train transitioning from one state to another.

[0108] Indicates that the observed value is observed in state i The probability of , represents the probability of observing the initial observation data in state i.

[0109] is a path pointer used to record the most likely previous state of state i at time t, and a pointer to record path information for path backtracking.

[0110] The termination condition can be expressed as:

[0111] It represents the maximum probability value of reaching a certain state at the end of the entire observation sequence (time T), which represents the maximum probability of the optimal path found by the Viterbi algorithm. It represents the maximum possibility of reaching a certain state through the optimal path at time T, and N represents the total number of moments.

[0112] is the endpoint state index of the optimal path, indicating the most likely state at time T, and indicating the state corresponding to time T in the optimal path calculated by the Viterbi algorithm. It is the state of the optimal path at the last moment.

[0113] It represents the probability of the most likely state sequence ending in state i at time T. It is the probability value of the optimal path to reach state i at time T. It represents the highest possibility that the train ends in state i after considering all possible preceding states.

[0114] T is the total length of the time series, indicating the last time point in the observation sequence. T is the moment when the observation sequence ends, which marks the final time point calculated by the Viterbi algorithm, so that the terminal state of the optimal path can be determined.

[0115] Through the above method, we can get the path backtracking formula:

[0116] Step 5: Optimization and decision making

[0117] The improved Newson algorithm takes into account the geometric properties, time constraints, topological structure and vehicle motion characteristics of the road network during the calculation process, and optimizes the selection of matching paths by dynamically adjusting the weights of state transition probabilities and observation probabilities.

[0118] In some embodiments, the method further comprises: Divide the travel route of the train into a plurality of initial segments, and determine the segmentation state of each initial segment; the segmentation state includes an observation state of the train passing through the initial segment; Based on the segment states of each initial segment and the corresponding state transition probability and state emission probability, the optimal path segment of the train's travel route is determined.

[0119] Specifically, the embodiment of the present application improves the heavy-haul railway path segmentation cutting method based on the hidden Markov model to improve the accuracy and efficiency of path analysis and management. The improved method includes the following key steps: First, the heavy-duty railway road (the road on which the train travels) is divided into sections of unequal length (initial segmentation) based on the actual terrain and path characteristics. The length of each section varies according to the environment. For example, in open areas, each section can be hundreds of meters long, while in complex areas such as tunnels, it can be shortened to tens of meters. Each section is assigned a corresponding status.

[0120] The segment state of the initial segment represents the geographical location and related attributes of a specific route segment of the heavy-haul railway. This includes the starting and ending locations of the route segment, mileage, road topology (such as bifurcations, intersections, etc.), and various state information (speed, acceleration, etc.) of the train passing through the route segment in a specific time period. In other words, the segment state includes the observed state of the train passing through the initial segment.

[0121] Through the Beidou positioning system, location points are marked at certain intervals in each section, which help to accurately map the state of the section. The state transition probability matrix of the hidden Markov model is improved to better simulate the continuity and smoothness of the heavy-haul railway path.

[0122] At the same time, the emission probability between each location point and the road segment state is considered to quantify the degree of match between them. These improvements enable more accurate cutting of heavy-load railway paths, as well as path segmentation and state inference.

[0123] The railway map matching method based on the Beidou navigation system and odometer provided in the embodiment of the present application solves the problem that the current mainstream map matching algorithm cannot be accurate to the section level, and optimizes the segmentation logic of heavy-duty railway lines. The improved hidden Markov model algorithm is used to segment heavy-duty railway lines more accurately, and the path segments are intelligently matched, which significantly improves the matching accuracy and overall performance of the heavy-duty railway system in complex scenarios.

[0124] In some embodiments, based on the segment states of each initial segment and the corresponding state transition probability and state emission probability, determining the optimal path segment of the train's travel route includes: Based on the Viterbi algorithm, the segment states of each initial segment and the corresponding state transition probability and state emission probability are recursively calculated to obtain the optimal path segmentation of the train's travel route.

[0125] Specifically, the method for constructing heavy-haul railway path segmentation based on the improved hidden Markov model and Viterbi algorithm is as follows: Step 1: Increase the density of location markers and mark location markers

[0126] Mark the position points at a certain interval in each paragraph, increase the density of position marking points, and determine the maximum length of each paragraph , define the state set .in, Indicates the state of the i-th paragraph. Each state Represents the geographic location and related attributes of a specific route segment of a heavy-haul railway. This includes the start and end locations of the route segment, mileage, road topology (such as bifurcations, intersections, etc.), and various status information (speed, acceleration, etc.) of the train passing through the route segment within a specific time period.

[0127] Mark a location mark on the road at a certain interval (for example, every 100 meters). Define the state transition probability matrix T, Indicates from the state Transfer to state probability.

[0128] Step 2: Add additional states

[0129] Define the state emission probability matrix E, Indicates status The probability of emitting observation j. There are sensors of the same type, but due to different functional domains, they have different priorities for data analysis.

[0130] Step 3: Path segmentation

[0131] The Viterbi algorithm is used to calculate the most likely hidden state sequence, that is, the most likely path segmentation result. The Viterbi algorithm selects the most likely state transition at each moment through dynamic programming to obtain the optimal path.

[0132] First initialize: Define the state collection: A state set represents different segments or sections of a heavy-haul railway path. In the map matching algorithm, each state si represents the state information of a train on a specific section. Specifically, si can contain the location information, speed, acceleration and other parameters that need to be tracked for that section. The state set defines all the sections that a train may be in, and tracks and matches the states of the train on these sections.

[0133] Define the observation sequence: An observation sequence represents a sequence of observations collected at different points in time. Each observation ot can include data obtained from Beidou, odometers, and other sensors, such as the vehicle's longitude, latitude, altitude, speed, and travel distance. These observations are inputs to the map matching process, and the actual trajectory and position of the train are inferred from these data.

[0134] Define the initial state probability vector: The initial state probability represents the probability that the train is in each state si at the initial moment (i.e. the moment when the train starts positioning). represents the initial probability of the train being in state si. This probability distribution defines the uncertainty of the train's state at the beginning, which helps the algorithm to infer the possible position of the train in the initial stage.

[0135] Define the state transition probability matrix: The state transition probability matrix represents the probability of transitioning from one state to another. Specifically, the elements in the matrix represents the probability of transitioning from state si to state sj. In heavy-haul railway systems, this matrix describes the probability of a train transitioning from one location or state to another. This matrix is ​​very important for trajectory tracking and path prediction because it defines the transition rules of the train between various states.

[0136] Define the state emission probability matrix: The emission probability matrix represents the probability of observing a particular observation value in a particular state. Specifically, the elements in the matrix represents the probability of observing the observation value oj in state si. This matrix reflects the relationship between states and observations. In map matching, the emission probability matrix is ​​used to estimate the likelihood of observing a given sensor data in a specific state (for example, a certain position and velocity).

[0137] Define the dynamic programming table: The dynamic programming table represents the maximum probability of being in a specific state at a specific time. Specifically, represents the probability of the most likely path ending in state si at the end of time t. This table is used in the recursive calculation of the Viterbi algorithm to determine the probability of the best path possible at each time point. This table combines the state transition probability and the emission probability to help find the most likely state sequence at each time point.

[0138] Recursive calculation: Initialize the dynamic programming table , record the product of the initial state probability and the state emission probability of the observation value o1.

[0139] For each state sj: calculate , record the maximum probability of state sj at the current moment.

[0140] Record path pointer Point to the state with the highest probability at the previous moment:

[0141] termination: Calculate the final probability

[0142] Path backtracking: Select the final state

[0143] By path pointer To backtrack: for : , update the optimal state at the current moment

[0144] Output: Get the state sequence of the optimal path segment: s1 represents the specific state of the train at time t = 1. This includes the train's initial position (longitude, latitude, altitude), initial speed, acceleration, and other relevant parameters.

[0145] s2 represents the state of the train at time t = 2. This includes the train's position information, speed, acceleration, etc. at the second time point.

[0146] sT represents the state of the train at time t = T. This includes the train's position information, speed, acceleration, etc. at the last point in time.

[0147] Each state si in the sequence S* represents the state at the corresponding time point, including position information and dynamic parameters, which together describe the motion trajectory of the train in the entire observation time series.

[0148] The railway map matching method based on the Beidou navigation system and the odometer provided in the embodiment of the present application divides the road into multiple sections according to actual needs and assigns an independent mapping state to each section, thereby reducing the adsorption error distance and improving the accuracy of map matching.

[0149] In some embodiments, the method further comprises: Analyze the trajectory data of the train and determine the angles between adjacent line segments based on the angles of adjacent line segments; Determine the distance between adjacent line segments based on the coordinates of the start points or end points of the adjacent line segments; Based on the slopes of the adjacent line segments, the degree of trajectory splitting of the adjacent line segments is determined; Based on the angle and distance between adjacent line segments, determine the penalty factor for merging intersections; Determine the penalty factor for non-merged intersections based on the angle, distance, and degree of trajectory splitting of adjacent line segments; Based on the merge intersection penalty factor and the non-merge intersection penalty factor, the selection probability of the intersection of adjacent line segments is determined.

[0150] Specifically, aiming at the optimization problem of selecting merging / non-merging intersections, the matching algorithm is improved to improve the path matching accuracy.

[0151] The innovation in handling merge / non-merge intersections is that different processing methods are used for merge and non-merge intersections to balance accuracy and the number of states processed by the algorithm. For merge intersections, the characteristic is that the states close to the intersection on different roads are merged into a state without specific road information. This processing method can reduce the number of states that the algorithm needs to process, but it sacrifices the accuracy of the matching results. Both the Graphhopper and Newson algorithms use the method of merging intersections. The advantage of this innovation is that it quickly reduces the number of states, thereby improving the efficiency of the algorithm. In contrast, the processing method for non-merge intersections requires each state to be directly linked to a specific road to reduce the possibility of incorrect road matching, especially in the presence of U-turn restrictions. Although this processing method increases the number of states that the algorithm needs to process, it can improve the accuracy of matching. By directly linking to specific roads, the road where each state is located can be more accurately determined, reducing the possibility of matching errors.

[0152] The embodiment of the present application takes into account the actual situation of heavy-haul railways and adopts a method of merging intersections to improve algorithm efficiency and reduce the number of states, including: Step 1: Trajectory Segmentation

[0153] Distance calculation between adjacent points: For adjacent line segments L1 and L2, point P1 on L1 ( , ) and P2 on L2( , ), calculate the distance between them , which can be expressed as:

[0154] Angle calculation between adjacent line segments: For adjacent line segments L1 and L2, calculate their angles at the starting point or end point of the line segment , which can be expressed as:

[0155] Step 2: Determine the intersection

[0156] By analyzing the trajectory data, the angle, distance and trajectory splitting degree between adjacent line segments are calculated.

[0157] Angle: ,in and are the angles between adjacent line segments L1 and L2 respectively.

[0158] distance: , where (x1, y1) and (x2, y2) are the start or end coordinates of adjacent line segments.

[0159] Track splitting degree: ,in and are the slopes of adjacent line segments L1 and L2 respectively.

[0160] Step 3: Introduce the merge / non-merge intersection penalty factor

[0161] In the process of path selection, a merging / non-merging intersection penalty factor is introduced to reduce the possibility of selecting these intersections.

[0162] Merge intersection penalty factor:

[0163] Non-merging intersection penalty factor:

[0164] in, , , , , is the weight coefficient of the penalty factor, , and is the threshold parameter.

[0165] The railway map matching method based on the Beidou navigation system and the odometer provided in the embodiment of the present application adopts a method of adopting a specific strategy when processing one-way section matching, thereby improving the accuracy of matching. When processing intersections, you can choose to merge or non-merge intersections to balance the accuracy of matching and the number of states processed according to requirements.

[0166] The system provided in the embodiment of the present application is described below. The system described below and the method described above can be referenced to each other.

[0167] Figure 2 This is a schematic diagram of the structure of the railway map matching system based on the Beidou navigation system and the odometer provided by this application. Figure 2 As shown, the system includes: The data acquisition module 210 is used to obtain the first state data of the train based on the Beidou navigation system, and to obtain the second state data of the train based on the odometer; A data fusion module 220 is used to fuse the first state data and the second state data to determine the observed state of the train at each time and the corresponding state transition probability and state emission probability; The map matching module 230 is used to determine the optimal matching path of the train at each time based on the observed state of the train at each time and the corresponding state transition probability and state emission probability.

[0168] The railway map matching system based on the Beidou navigation system and the odometer provided in the embodiment of the present application obtains the first state data of the train based on the Beidou navigation system, and obtains the second state data of the train based on the odometer; the first state data and the second state data are fused and processed to determine the observed state of the train at each time and the corresponding state transition probability and state emission probability; based on the observed state of the train at each time and the corresponding state transition probability and state emission probability, the optimal matching path of the train at each time is determined; by efficiently fusing and analyzing the multi-source data of the Beidou navigation system and the odometer, it is ensured that these data are fully utilized and optimized in the map matching process; according to the observed state of the train at each time and the corresponding state transition probability and state emission probability, the optimal matching path of the train at each time is determined, and the adaptability of map matching in different heavy-load railway network structures and changing environmental conditions is enhanced through improved algorithm design, ensuring that high-precision and high-reliability positioning services can be provided even when the Beidou signal is weak or fails; the positioning accuracy and stability of heavy-load railway vehicles in complex environments are improved.

[0169] Figure 3 This is a schematic diagram of the overall solution of the railway map matching method based on the Beidou navigation system and odometer provided by this application, such as Figure 3 As shown, the method as a whole may include: (1) Data collection and preliminary processing: involves the collection and preliminary processing of Beidou and odometer data. In this stage, Beidou and odometer data are efficiently collected, covering positioning data, driving distance and vehicle dynamic information. These data are initially filtered and fused, especially the Kalman filter is used to accurately fuse the odometer and Beidou data to reduce noise and obvious errors in the data, while filtering out useful mileage information. In addition, the hidden Markov model optimizes the processing of state transition and observation probability in the map matching process, combined with the Viterbi algorithm, to ensure that all types of data are fully utilized and optimized in the map matching process, thereby improving the overall matching accuracy and system efficiency.

[0170] (2) Map matching and state management: The improved Newson algorithm is used for accurate map matching. In this stage, detailed information after data preprocessing is used to perform detailed state discretization and segmentation, and advanced state management functions are introduced to accurately adapt to complex environments and vehicle dynamics, optimize the map matching process, and effectively reduce mismatches. At the same time, by finely controlling the generation of the state space, its infinite expansion is avoided, thereby improving the processing efficiency of the algorithm.

[0171] (3) Map update: Use the fused and processed data to revise and update the map to ensure the timeliness and accuracy of the map data. This process demonstrates a systematic approach that efficiently integrates and processes multi-source data through the application of advanced algorithms and technologies to support accurate and reliable heavy-duty railway positioning and navigation systems.

[0172] This method ensures a seamless transition from efficient data collection to preliminary processing from multi-source sensor data by tightly integrating data collection and preliminary processing, providing accurate and optimized input data for the map matching algorithm. In the map matching and state management stage, the improved Newson algorithm significantly improves the matching accuracy and enhances the algorithm's adaptability to complex traffic environments through fine state discretization and segmentation, as well as advanced state management functions. In addition, the introduction of advanced feature processing steps further optimizes the matching process, effectively copes with U-turns, circular routes and other complex conditions in heavy-duty railway networks, and significantly reduces the possibility of mismatching.

[0173] Figure 4 This is a data fusion diagram of the railway map matching method based on the Beidou navigation system and odometer provided in this application, such as Figure 4 As shown, including: (1) The data collection and preprocessing step mainly focuses on collecting raw data from the Beidou system and odometer. At this stage, all collected data are first fused through the Kalman filter. The Kalman filter plays a vital role here. It can effectively integrate the data from Beidou and odometer, and optimize the estimation of position and motion state by considering the error characteristics of various sensors. In addition, data standardization, time synchronization and preliminary filtering processing, such as removing noise and outliers, are also performed to ensure the accuracy and consistency of the data. After the preliminary state estimation and fusion, these preprocessed data can be transmitted to the next step, which is the data processing and analysis stage.

[0174] (2) The data processing and analysis steps are mainly carried out through hidden Markov models (HMM) and Viterbi algorithms. First, HMM is used to establish a relationship model between sensor data and possible states to predict various possible vehicle position states. Subsequently, the Viterbi algorithm uses these predictions and refines them through its dynamic programming characteristics to determine the most likely state sequence, thereby achieving accurate calculation of the optimal path. This calculation process takes into account information extracted from multi-source heterogeneous data (such as Beidou and odometer data) and optimizes the map matching process. This not only helps to reduce false matches, but also improves the overall accuracy of map matching. Through this series of processing and analysis strategies, the effective management and optimized use of data are ensured, greatly improving the efficiency and accuracy of the entire data processing process.

[0175] Figure 5 is a schematic diagram of the structure of the electronic device provided by this application, such as Figure 5 As shown, the electronic device may include: a processor (Processor) 510, a communication interface (Communications Interface) 520, a memory (Memory) 530 and a communication bus (Communications Bus) 540, wherein the processor 510, the communication interface 520, and the memory 530 communicate with each other through the communication bus 540. The processor 510 may call the logic command in the memory 530 to execute the method described in the above embodiment, for example: The first state data of the train is obtained based on the Beidou navigation system, and the second state data of the train is obtained based on the odometer; the first state data and the second state data are fused to determine the observed state of the train at each moment and the corresponding state transition probability and state emission probability; based on the observed state of the train at each moment and the corresponding state transition probability and state emission probability, the optimal matching path of the train at each moment is determined.

[0176] In addition, the logical commands in the above-mentioned memory can be implemented in the form of software functional units and can be stored in a computer-readable storage medium when sold or used as an independent product. Based on this understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several commands to enable a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc., which can store program code.

[0177] The processor in the electronic device provided in the embodiment of the present application can call the logic instructions in the memory to implement the above method. Its specific implementation method is consistent with the implementation method of the aforementioned method and can achieve the same beneficial effects, which will not be repeated here.

[0178] An embodiment of the present application also provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the methods provided in the above embodiments are implemented.

[0179] Its specific implementation is consistent with the aforementioned method implementation and can achieve the same beneficial effects, so it will not be repeated here.

[0180] An embodiment of the present application provides a computer program product, including a computer program, which implements the above method when executed by a processor.

[0181] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0182] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0183] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A railway map matching method based on Beidou navigation system and odometer, characterized in that: include: Acquiring first status data of a train based on a Beidou navigation system, and acquiring second status data of the train based on an odometer; Performing fusion processing on the first state data and the second state data to determine the observed state of the train at each time and the corresponding state transition probability and state emission probability; Based on the observed state of the train at each moment and the corresponding state transition probability and state emission probability, the optimal matching path of the train at each moment is determined.

2. The railway map matching method based on Beidou navigation system and odometer according to claim 1, characterized in that: The fusing the first state data and the second state data to determine the observed state of the train at each time and the corresponding state transition probability and state emission probability includes: Performing fusion processing on the first state data and the second state data based on a Kalman filter to obtain state fusion data; The state fusion data is processed based on a hidden Markov model and a Viterbi algorithm to obtain the observed state of the train at each moment and the corresponding state transition probability and state emission probability.

3. The railway map matching method based on Beidou navigation system and odometer according to claim 2 is characterized in that: Before fusing the first state data and the second state data based on the Kalman filter to obtain state fusion data, the method further includes: The first state data and the second state data are preprocessed; the preprocessing includes at least one of data standardization, time synchronization and filtering processing.

4. The railway map matching method based on Beidou navigation system and odometer according to claim 2 is characterized in that: The state fusion data is processed based on the hidden Markov model and the Viterbi algorithm to obtain the observed state of the train at each moment and the corresponding state transition probability and state emission probability, including: Processing the state fusion data based on a hidden Markov model, establishing a relationship between the state fusion data and the possible state of the train, and determining the possible state of the train at each moment and the corresponding state transition probability and state emission probability; Based on the Viterbi algorithm, possible states of the train at each moment and corresponding state transition probabilities and state emission probabilities are recursively calculated to determine the observed state of the train at each moment.

5. The railway map matching method based on Beidou navigation system and odometer according to claim 1, characterized in that: The method further comprises: Dividing the travel route of the train into a plurality of initial segments, and determining the segmentation state of each initial segment; the segmentation state includes an observed state of the train passing through the initial segment; Based on the segment states of the initial segments and the corresponding state transition probabilities and state emission probabilities, the optimal path segment of the travel road of the train is determined.

6. The railway map matching method based on Beidou navigation system and odometer according to claim 5 is characterized in that: The determining of the optimal path segment of the travel road of the train based on the segment states of each initial segment and the corresponding state transition probability and state emission probability comprises: Based on the Viterbi algorithm, the segment states of each initial segment and the corresponding state transition probability and state emission probability are recursively calculated to obtain the optimal path segment of the travel road of the train.

7. The railway map matching method based on the Beidou navigation system and odometer according to any one of claims 1 to 6, characterized in that: The method further comprises: Analyzing the trajectory data of the train, and determining the angle between adjacent line segments based on the angles of adjacent line segments; Determining the distance between the adjacent line segments based on the coordinates of the starting points or the end points of the adjacent line segments; Determining a degree of trajectory splitting of the adjacent line segments based on the slopes of the adjacent line segments; Determine a penalty factor for merging intersection points based on the angle and distance between the adjacent line segments; Determining a non-merged intersection penalty factor based on the angle, distance and trajectory splitting degree of the adjacent line segments; Based on the merged intersection penalty factor and the non-merged intersection penalty factor, the selection probability of the intersection of the adjacent line segments is determined.

8. A railway map matching system based on Beidou navigation system and odometer, characterized in that: include: A data acquisition module, used for acquiring first status data of a train based on a Beidou navigation system, and acquiring second status data of the train based on an odometer; A data fusion module, used for fusing the first state data and the second state data to determine the observed state of the train at each time and the corresponding state transition probability and state emission probability; The map matching module is used to determine the optimal matching path of the train at each moment based on the observed state of the train at each moment and the corresponding state transition probability and state emission probability.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the railway map matching method based on the Beidou navigation system and the odometer as described in any one of claims 1 to 7 is implemented.

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the railway map matching method based on the Beidou navigation system and the odometer as described in any one of claims 1 to 7 is implemented.

Images