A train integrity status acquisition method based on optimized fuzzy reasoning

Abstract

The present invention provides a method for obtaining train integrity status based on optimized fuzzy reasoning. The method includes: collecting satellite positioning decision information of the front and rear ends of the target train, train running speed and acceleration information, and constructing a basic feature data set of the train integrity status; collecting the paired relative distances between the locomotives and each carriage contained in the train, and constructing an enhanced feature data set of the train integrity status; establishing a basic fuzzy reasoning system for train integrity status decision, determining a fuzzy rule table and defuzzification rules; updating the membership function parameters of the basic fuzzy reasoning system based on intelligent optimization decision logic, implementing integrity event alarms based on the integrity status decision results output by the optimized fuzzy reasoning system, and determining the integrity associated area boundaries based on the front and rear positions of the train. The present invention can realize accurate and rapid train integrity status decision-making, and ensure timely alarms for train integrity events under complex operating conditions and scenarios.

Description

Technical Field

[0001] The present invention relates to the technical field of train status detection, and in particular to a method for acquiring train integrity status based on optimized fuzzy reasoning. Background Art

[0002] During the operation of railway trains, due to factors such as line conditions and special circumstances, the driver may need to brake or accelerate the train, so that each carriage moves in a different state within a short time and short distance, causing friction between the couplers connecting two adjacent carriages, which will affect the connection quality between the carriages. Under long-term and high-frequency wear, if the safety status of the couplers is not detected and maintained in time, and the hidden dangers of broken or unhooked couplers are reduced or eliminated, once the train has integrity events such as unhooking / broken couplers, the separated carriages will gradually slow down due to loss of power until they stay in the line section. If the following car does not find the stranded car in front, it may cause operational safety risks.

[0003] Train integrity refers to an important factor directly related to the structural status of the train and the driving safety risks of the preceding and following vehicles according to the requirements of marshaling and transportation, indicating the complete status of the overall connectivity of the locomotive and carriages of the entire train during the driving process. The detection and decision of the train integrity status is generally completed by the Train Integrity Monitoring System (TIMS). TIMS is an important subsystem for realizing the train integrity detection function in the train operation control system. By real-time detection, confirmation and response to the train integrity status, the driver and the ground control center can effectively identify the train operation status in real time and guide safe disposal operations in the event of a train integrity event. Therefore, timely and accurate acquisition of the train integrity status is of great significance to the safe and efficient operation of the train.

[0004] At present, a train integrity status detection method in the prior art includes: relying on the tail device to monitor the train air duct pressure in real time, and judging whether it is necessary to alarm the driver according to whether the air duct pressure value is normal or abnormal. The driver can take different actions corresponding to the situation according to the wind pressure status prompts of the control device.

[0005] The disadvantages of the above-mentioned train integrity status detection method in the prior art include: the method based on duct pressure monitoring has continuity and availability risks, and its response capability to integrity events and the existence of false alarms require the introduction of additional technical means to compensate or replace it in practical applications. In addition, with the continuous development of train operation management and control systems, the technical idea with "on-board autonomy" as the core further emphasizes the great value of the train's own detection and processing capabilities.

[0006] At present, another train integrity status detection method in the prior art includes: implementing vehicle length estimation based on satellite positioning, and determining whether a train separation event has occurred by calculating the longitudinal distance between the front and rear ends of the train and comparing it with the reference vehicle length. Representative work includes: obtaining a more accurate position estimate based on the satellite positioning double difference estimate of the front and rear ends of the train, thereby achieving accurate vehicle length estimation and implementing vehicle length comparison and logical judgment, or combining the strategy of implementing vehicle length estimation and judgment based on the front and rear end positioning of the train with traditional wind pressure monitoring to further compensate and enhance the solution that only uses wind pressure monitoring.

[0007] Another train integrity status detection method in the prior art has the following disadvantages: the estimation and monitoring of train length based on satellite positioning requires good signal observation conditions in the operating environment of the train. However, many railway lines have closed and shielded environments such as tunnels and stations. In these operating scenarios, satellite positioning cannot obtain effective signal observation, and thus cannot complete the expected positioning, train length estimation and other operations, resulting in the expected intention to enhance the wind pressure monitoring method. In addition, during the operation of the train, there will be certain performance degradation problems in satellite positioning due to the surrounding terrain environment, signal interference conditions, etc., which may cause possible inaccuracy risks in satellite positioning-driven train length estimation, affecting its effective integrity status decision-making and response. Summary of the invention

[0008] The embodiment of the present invention provides a method for obtaining the integrity status of a train based on optimized fuzzy reasoning, so as to effectively determine the integrity-related area boundary of the train and make accurate and rapid decisions on the integrity status of the train.

[0009] In order to achieve the above object, the present invention adopts the following technical scheme.

[0010] A train integrity status acquisition method based on optimized fuzzy reasoning, comprising:

[0011] The front-end satellite positioning decision information, train running speed and acceleration information of the target train are collected through the on-board equipment at the front end of the train, and the rear-end satellite positioning decision information, train running speed and acceleration information are collected through the on-board equipment at the rear end of the train. Combined with the track space database of the target line, the basic characteristic data set of the train integrity status is constructed;

[0012] Collecting the pairwise relative distances between the locomotives and the carriages contained in the train through an ultra-wideband sensor, associating with the train identity through time alignment, and constructing a train integrity status enhanced feature data set based on the train integrity status basic feature data set according to the pairwise relative distances between the locomotives and the carriages contained in the train;

[0013] Establishing a basic fuzzy reasoning system for train integrity status decision-making based on the train integrity status enhanced feature data set, and determining a fuzzy rule table and a defuzzification rule of the basic fuzzy reasoning system;

[0014] Based on the intelligent optimization decision logic, the membership function parameters of the basic fuzzy inference system are updated to establish an optimized fuzzy inference system for train integrity status decision;

[0015] An integrity event alarm is implemented based on the integrity status decision result output by the optimized fuzzy inference system, and the integrity associated area boundary of the train is determined based on the position calculation of the front end and the rear end of the train.

[0016] Preferably, the front-end satellite positioning decision information, train running speed and acceleration information of the target train are collected by the on-board equipment at the front end of the train, the rear-end satellite positioning decision information, train running speed and acceleration information are collected by the on-board equipment at the rear end of the train, and the basic characteristic data set of the train integrity status is constructed in combination with the target line track space database, including:

[0017] The on-board equipment at the front end of the train uses the satellite positioning terminal contained in it to collect satellite positioning decision information at the roof antenna at the front end of the train. The satellite positioning decision information includes the longitude of the roof antenna at the front end of the train at time t. latitude and elevation The inertial measurement terminal is used to collect and calculate the running status information of the front end of the train, which includes the acceleration of the front end of the train at time t. and speed

[0018] The on-board equipment at the rear of the train uses the satellite positioning terminal contained in it to collect satellite positioning decision information at the built-in antenna at the rear end of the train. The satellite positioning decision information includes the longitude of the rear end of the train at time t. latitude and elevation Use the inertial measurement terminal to collect and calculate the running status information of the rear end of the train, which includes the acceleration of the rear end of the train at time t. and speed

[0019] Based on the coordinate position of the antenna on the roof of the front end of the train Rear end of train Track frame position at the previous moment Based on the search scale coefficient η, the candidate track feature point subsets of the front and rear ends of the train are retrieved and extracted forward along the train running direction.

[0020] Among them, Pi x Indicates the i-th track feature point at the x end within the search scale coefficient range, based on which the minimum relevant segment in the track frame corresponding to the target position is determined Solve the front and rear track frame positions of the train at the current moment Combined with the train front roof antenna registration parameter Δ H , the internal antenna alignment parameters of the rear end of the train Δ E Calculate the estimated train length:

[0021]

[0022] (4) Construct the basic characteristic data set of the train integrity status at time t based on the front and rear track frame positions, running speed, acceleration and train length estimation information of the target train

[0023] Preferably, the collecting of the paired relative distances between the locomotives and the carriages contained in the train by ultra-wideband sensors, associating with the train identity by time alignment, and constructing the train integrity status enhanced feature data set based on the train integrity status basic feature data set according to the paired relative distances between the locomotives and the carriages contained in the train include:

[0024] Ultra-wideband sensors are installed in the locomotives and carriages inside the train to collect the relative distances between pairs of carriages in real time. According to the distribution of locomotives and carriages, it is assumed that the train contains N carriages. The nearest locomotive carriage is numbered CAR(1), and the carriages are numbered successively to CAR(N) in the direction away from the locomotive. There are two cases for distance measurement information observation:

[0025] 1) Train locomotive, the distance D between the train locomotive and the adjacent carriage connected to it is measured by the ultra-wideband sensor installed on the rear edge of the locomotive LOC,t , where the measured pairwise relative distances include the offset δ between the locomotive ultra-wideband sensor and the rear edge of the locomotive loc , Carriage CAR (1) ultra-wideband sensor offset from the front edge of the car Distance between locomotive and car (1) d LOC-CAR(1),t ,Right now

[0026]

[0027] For the locomotive, an ultra-wideband sensor is installed in the direction of the adjacent car CAR (1), and the pre-calibrated δ loc For the carriage CAR (1), an ultra-wideband sensor is installed in the direction of the adjacent locomotive and the adjacent carriage CAR (2). LOC-CAR(1),t Related constraint threshold judgment and reasoning analysis, pre-calibration Calibration parameters in the direction of adjacent carriage CAR(2) The calibration of will be used for D LOC,t Related constraint boundary threshold judgment and reasoning analysis.

[0028] 2) For train carriage CAR(i), where i < N, the distance D between carriage CAR(i) and its adjacent carriage CAR(i + 1) is measured by an ultra-wideband sensor installed at the rear edge of the carriage CAR(i),t , where the measured pairwise relative distances include the offset of the ultra-wideband sensor at the rear end of carriage CAR(i) from the rear edge of the carriage The offset of the ultra-wideband sensor at the front end of carriage CAR(i + 1) from the front edge of the carriage The distance d between carriages CAR(i) and CAR(i + 1) CAR(i)-CAR(i+1),t , that is

[0029]

[0030] Among them, for carriage CAR(i), one ultra-wideband sensor is installed in the direction of adjacent carriage CAR(i - 1) and adjacent carriage CAR(i + 1) respectively, and pre-calibrated Among them The calibration of will be used for D CAR(i-1),t Related constraint boundary threshold judgment and reasoning analysis The calibration of will be used for D CAR(i),t Related constraint boundary threshold judgment and reasoning analysis;

[0031] 3) For train carriage CAR(N), the distance D between carriage CAR(N - 1) and CAR(N) is measured by an ultra-wideband sensor installed at the rear edge of carriage CAR(N - 1) CAR(N-1),t , for carriage CAR(N), one ultra-wideband sensor is installed in the direction of adjacent carriage CAR(N - 1), and pre-calibrated Used for D CAR(N-1),t Related constraint boundary threshold judgment and reasoning analysis;

[0032] (2) After the train integrity detection management platform obtains the pairwise relative distance information measured by the wireless sensors sent by the locomotive and carriages, it performs time alignment and train identity association on the relative distance information through time index and identity index respectively, where:

[0033] Time alignment means that for the timestamp information in the set of relative distance information, the information in the same time and aligned with the train integrity status decision processing period T det is grouped into clusters, taking t = t p 、t = t p +T det as an example, the obtained cluster is

[0034]

[0035] in, t p Time information cluster, Indicates that the ID is n j The distance between the locomotive of a train and the adjacent carriage connected to it is t p The measured distance at a moment, Indicates that the ID is n k The distance between the locomotive of a train and the adjacent carriage connected to it is t p The measured distance at a moment, Indicates that the ID is n x The distance between the i-th car of the train and the adjacent (i+1)-th car connected to it is t p The measured distance at a moment, Indicates that the ID is n y The distance between the u-th car of the train and the adjacent (u+1)-th car connected to it is t p The measured distance at a moment.

[0036] Train identity association refers to the secondary segmentation of the clusters aligned by time using the train ID number as the index to form an ultra-wideband measurement relative distance data set indexed by a specific time and a specific train identity. p , t=t p +T det And the train ID number is n j For example, the relative distance data set obtained by observing ultra-wideband ranging information is:

[0037]

[0038] in, Indicates t p The time is based on the train ID number n j is the relative distance dataset obtained by indexing, It means (t p +T det ) time with train ID number n j is the relative distance dataset obtained by index, N(n j ) indicates that the ID number is n j The total number of carriages in a train excluding locomotives, Indicates that the ID is n j The distance between the locomotive of a train and the adjacent carriage connected to it is t p The measured distance at a moment, Indicates that the ID is n j The distance between the first carriage of the train and the second adjacent carriage connected to it is t pThe measured distance at a moment, Indicates that the ID is n j The train number (N(n j )-1) car and the adjacent N(n)th car connected to it j ) Between the carriages at t p The measured distance at a moment, Indicates that the ID is n j The distance between the locomotive of the train and the adjacent carriage connected to it is (t p +T det ) at the time of measurement, Indicates that the ID is n j The distance between the first carriage of the train and the second adjacent carriage connected to it is (t p +T det ) at the time of measurement, Indicates that the ID is n j The train number (N(n j )-1) car and the adjacent N(n)th car connected to it j ) carriages between (t p +T det ) at the moment of measurement;

[0039] (3) Using time information and train ID as index, The corresponding ID is n j The basic characteristic data set of train integrity status Ψ is calculated by train collection t (n j ) to form a train integrity status enhanced feature dataset The data information items it contains are further expanded as follows:

[0040]

[0041] Among them, D loc,t (n j ) indicates that the ID is n j The measured distance between the locomotive of a train and the adjacent carriage connected to it at time t, D CAR(1),t (n j ) indicates that the ID is n j The measured distance between the first carriage of the train and the second adjacent carriage connected to it at time t, Indicates that the ID is n j The train number (N(n j )-1) car and the adjacent N(n)th car connected to it j ) is the measured distance between the carriages at time t, Indicates that the ID is n jThe front track frame position of the train at time t, Indicates that the ID is n j The front running speed of the train at time t, Indicates that the ID is n j The front acceleration of the train at time t is Indicates that the ID is n j The rear track frame position of the train at time t, Indicates that the ID is n j The rear running speed of the train at time t, Indicates that the ID is n j The rear end acceleration of the train at time t is Indicates that the ID is n j The estimated length of the train at time t.

[0042] Preferably, the establishing of a basic fuzzy inference system for train integrity status decision-making based on the train integrity status enhanced feature data set and the determination of a fuzzy rule table and a defuzzification rule of the basic fuzzy inference system include:

[0043] Based on the various information items contained in the enhanced feature data set of train integrity status, a basic fuzzy inference system for train integrity status decision-making is established. The basic fuzzy inference system for train integrity status decision-making is used to determine the specific time t and specific train n. j Implement integrity status determination;

[0044] In the process of establishing the basic fuzzy inference system for train integrity status decision, a four-input single-output system is designed. Based on the definition of fuzzy subsets, each fuzzy input has three cases: normal SN, possibly normal SM and abnormal SU, and the fuzzy output has two cases: complete TN and loss of integrity TU. The corresponding fuzzy rule table is established, which contains 81 fuzzy rules in total, as shown in Table 1:

[0045] Table 1 Fuzzy rules of basic fuzzy inference system for train integrity status decision

[0046]

[0047]

[0048] In the process of establishing the basic fuzzy inference system for train integrity status decision, the state judgment output module implements the defuzzification operation using the defuzzification rules to activate Fuzzy rules are merged, and the output of each rule A single set of fuzzy rules aggregated into the output variable is obtained using the maximum operator notation:

[0049]

[0050] The output of the fuzzy rule set calculated using the aggregation method is:

[0051]

[0052] Based on the membership function, specific fuzzy rules are established, and the fuzzy rule input is mapped to the fuzzy output. Through the steps of fuzzification, fuzzy reasoning and defuzzification, the input information is converted into a fuzzy description of the output to realize fuzzy reasoning. Based on the fuzzy description of the output obtained by fuzzy reasoning operation, the fuzzy output is processed by defuzzification operation in the state judgment output module to obtain the train integrity state decision result as a complete TN state or a lost integrity TU state.

[0053] Preferably, the establishment process of the basic fuzzy inference system for train integrity status decision-making includes:

[0054] The various types of information contained in the train integrity status enhanced feature data set are grouped into four information items: train length information item Train front and rear speed deviation information items Train front and rear end acceleration deviation information items and the pairwise relative distance information item D between carriages γ,t (n j ), the four information items are parameterized by difference and their ranges are constrained. The corresponding parameter restriction rules are as follows:

[0055]

[0056] in, is the nominal length of the train, ΔL, ΔV, ΔA, and ΔD are the constraints of the train length deviation, the speed deviation of the front and rear ends of the train, the acceleration deviation of the front and rear ends of the train, and the relative distance between the pairs of carriages, respectively;

[0057] Based on the information item after standardized transformation as input, the triangular membership function is used as the membership function of the basic fuzzy inference system for train integrity status decision, which is used to describe the degree of belonging of each input information item element to the fuzzy set. The expression of the triangular membership function corresponding to the three cases of normal SN, possible normal SM and abnormal SU is:

[0058]

[0059] Among them, f1 * (x), Respectively D γ,t The membership function of the four information items, {p k}, k = 1 to 4 are the right end points of the four SN triangle membership functions, {a k ,b k}, k = 1 to 4 are the left starting point, vertex, and right ending point of the four SM triangle membership functions, {q k}, k=1~4 are the left starting points of the four SU triangular membership functions, and F is the horizontal coordinate boundary value of the membership function.

[0060] Preferably, the method of updating the membership function parameters of the basic fuzzy inference system based on the intelligent optimization decision logic to establish an optimized fuzzy inference system for train integrity status decision-making includes:

[0061] The intelligent optimization decision logic is used to implement the iterative optimization determination of the membership function parameters of the basic fuzzy inference system. The execution of the intelligent optimization decision logic specifically includes the following steps:

[0062] Step 4.1: Initialize the population to be optimized

[0063] The population to be optimized consists of a particle set Composition, wherein each particle represents a set of triangular membership function parameter sets, including control point parameters of each fuzzy set membership function;

[0064] Step 4.2: Particle fitness evaluation

[0065] For each particle in the population, the corresponding fuzzy inference system file is generated according to the membership function parameters represented by the particle. The acquired train integrity status enhanced feature data set is tested on the generated fuzzy inference system, and the obtained accuracy value is used as the fitness measure of each particle for feedback. The fitness is expressed as:

[0066]

[0067] Among them, TP represents true positive examples, that is, the number of positive samples correctly predicted by the system as positive; TN represents true negative examples, that is, the number of negative samples correctly predicted by the system as negative; FP represents false positive examples, that is, the number of negative samples incorrectly predicted by the system as positive; FN represents false negative examples, that is, the number of positive samples incorrectly predicted by the system as negative;

[0068] Step 4.3: Individual and global optimal updates

[0069] For each particle, first compare its current fitness with its individual historical optimal fitness. If the current fitness is better, update the individual optimal solution of the particle to further explore the potential optimal solution space. Secondly, for the entire particle set, search for the particle with the best fitness and take its fitness value as the global optimal solution. Finally, by updating the global optimal solution, guide the entire population to move towards a better solution.

[0070] Step 4.4: Particle velocity and position update

[0071] According to the individual optimal solution of each particle and the global optimal solution, the particle evolution update speed and position are updated. The speed update takes into account the particle's current speed, the direction from the current position to the individual optimal position, and the direction from the current position to the global optimal position;

[0072] Step 4.5: Iteration termination condition detection

[0073] Check whether the population update iteration reaches the iteration termination condition, which includes two types:

[0074] (1) The current population iteration number has reached the maximum number of iterations I max ;

[0075] (2) The global optimal solution obtained in the current iteration step is better than the preset threshold W max .

[0076] When any one of the above two conditions is met, it indicates that the optimization iteration of the membership function parameters required to construct the train integrity status decision optimization fuzzy inference system has been completed, and the subsequent steps will be entered to complete the update and construction of the train integrity status decision optimization fuzzy inference system; when both of the above two conditions are not met, return to step 4.2 and continue the next round of iteration;

[0077] Step 4.6: Confirm the optimization parameters of the membership function

[0078] Whether the value of the membership parameter corresponding to the global optimal solution of the current iteration step meets the range condition, if the range constraint condition is met, the group of global optimal solutions is used as the output to confirm the optimal set of membership function parameters required for the train integrity status decision optimization fuzzy inference system;

[0079] The optimal set of membership function parameters obtained by iterative optimization of intelligent optimization decision logic is used to update the basic fuzzy reasoning system for train integrity status decision, and the triangular membership functions corresponding to the three situations of normal (SN), possibly normal (SM), and abnormal (SU) in the fuzzy reasoning operation module are optimized. The fuzzy reasoning is implemented using the fuzzy rule table shown in Table 1, and the fuzzy output is processed through defuzzification operation in the state judgment output module to obtain a final and accurate integrity status decision result.

[0080] Preferably, the implementation of integrity event alarm based on the integrity status decision result output by the optimized fuzzy inference system, and determining the integrity associated area boundary of the train based on the position of the front end and the rear end of the train, includes:

[0081] (1) The real-time data collected during train operation is used to update the enhanced feature data set of the train integrity status. The optimized fuzzy inference system for train integrity status decision-making implements reasoning and decision-making. Based on the integrity status decision results output by the system, the following response operations are performed in the event of an integrity event:

[0082] 1) When the train integrity status decision optimization fuzzy inference system outputs the ID at time t, it should be n j When the integrity status of the train is complete TN, train n j The integrity event flag remains F TIM =0;

[0083] 2) When the train integrity status decision optimization fuzzy inference system outputs the ID at time t, it should be n j When the integrity state of the train is lost integrity TU, train n j The integrity event flag F TIM Set to 1 and lock F TIM Value modification rights, based on F TIM =1 This condition generates integrity event alarm information and outputs it;

[0084] (2) When the optimized fuzzy inference system of the train integrity status decision implements the reasoning decision and determines the integrity status output, the corresponding integrity association area boundary is generated synchronously to assist in confirming the train's track occupancy when the train integrity status is in the complete TN state, and to assist in confirming the integrity association area boundary range that needs to be protected when the front and rear ends of the train are separated when the train integrity status is in the lost integrity TU state. The specific principles are as follows:

[0085] 1) When the train integrity status decision optimization fuzzy inference system outputs the ID at time t, it should be n j When the integrity status of the train is complete TN, the front and rear ends of the integrity association area boundary are calculated based on the estimated solution of the front position of the train and the length of the train. According to the availability conditions of the satellite positioning of the front and rear ends of the train, it is determined in two cases:

[0086] Case 1: When satellite positioning is available

[0087] Integrity Association Region Boundary Front End for:

[0088]

[0089] in, The track frame position calculated for the train front onboard equipment, is the front-end margin, and its sign is determined by the up and down directions of the train.

[0090] Integrity Association Region Boundary Backend for:

[0091]

[0092] in, The calculated track frame position for the rear-end onboard equipment, is the rear-end margin, and its sign is determined by the up and down directions of the train.

[0093] Case 2: When satellite positioning is not available

[0094] Integrity Association Region Boundary Front End for:

[0095]

[0096] in, The train front position information is extracted from the speed and distance measurement module contained in the train control on-board equipment.

[0097] For The corresponding front-end margin, the sign is determined by the up and down direction of the train;

[0098] Integrity Association Region Boundary Backend for:

[0099]

[0100] 2) When the train integrity status decision optimization fuzzy inference system outputs the ID at time t, it should be n j When the integrity status of the train is loss of integrity (TU), the front and rear ends of the integrity association area boundary are calculated based on the estimated solution of the front and rear ends of the train under the most unfavorable conditions. According to the availability conditions of satellite positioning at the front and rear ends of the train, it is determined in two cases:

[0101] Case 1: When satellite positioning is available

[0102] Integrity Association Region Boundary Front End for:

[0103]

[0104] Integrity Association Region Boundary Backend for:

[0105]

[0106] Case 2: When satellite positioning is not available

[0107] Integrity Association Region Boundary Front End for:

[0108]

[0109] Integrity Association Region Boundary Backend for:

[0110]

[0111] in, For the train at t TU0 From the moment n TU0 Function of the distance traveled from the disconnection of the first carriage to the disconnection of the rear end of the train at the current time t, m i is the mass of the i-th carriage, χ m The most unfavorable slope condition for the target line.

[0112] It can be seen from the technical solutions provided by the above-mentioned embodiments of the present invention that the present invention proposes a method and system that can effectively utilize train-mounted detection information to implement multi-source fusion integrity status decision-making, providing a more solid implementation approach for accurate, fast and effective judgment and decision-making on the integrity status of train in-transit operation.

[0113] Additional aspects and advantages of the present invention will be given in part in the following description, which will become obvious from the following description, or may be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0114] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0115] Figure 1 A processing flow chart of a train integrity status acquisition method based on optimized fuzzy reasoning provided by an embodiment of the present invention;

[0116] Figure 2 A schematic diagram of a train length estimation principle based on the front and rear end positions of a train provided in an embodiment of the present invention.

[0117] Figure 3A schematic diagram of the relative distance measurement principle between a locomotive / carriage and between carriages / carriages provided in an embodiment of the present invention.

[0118] Figure 4 A processing flow chart for updating membership function parameters based on intelligent optimization decision logic is provided in an embodiment of the present invention.

[0119] Figure 5 A schematic diagram for comparing train integrity status decision results provided by an embodiment of the present invention.

[0120] Figure 6 A schematic diagram of an implementation principle of a train integrity status decision system based on optimized fuzzy reasoning provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0121] The embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and cannot be interpreted as limiting the present invention.

[0122] It will be understood by those skilled in the art that, unless expressly stated, the singular forms "one", "said", and "the" used herein may also include plural forms. It should be further understood that the term "comprising" used in the specification of the present invention refers to the presence of the features, integers, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or groups thereof. It should be understood that when we refer to an element as being "connected" or "coupled" to another element, it may be directly connected or coupled to the other element, or there may be intermediate elements. In addition, the "connection" or "coupling" used herein may include wireless connection or coupling. The term "and / or" used herein includes any unit and all combinations of one or more associated listed items.

[0123] It will be understood by those skilled in the art that, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as those generally understood by those skilled in the art in the art to which the present invention belongs. It should also be understood that terms such as those defined in common dictionaries should be understood to have meanings consistent with the meanings in the context of the prior art, and will not be interpreted with idealized or overly formal meanings unless defined as herein.

[0124] To facilitate understanding of the embodiments of the present invention, several specific embodiments will be further explained below with reference to the accompanying drawings, and each embodiment does not constitute a limitation on the embodiments of the present invention.

[0125] The present invention is based on the key role of train integrity detection in ensuring the safety of train operation. Taking into account the availability issues and performance risk points faced by existing technical solutions and ideas, a method and system are proposed that can effectively utilize train on-board detection information to implement multi-source fusion integrity status decision-making, providing a more solid implementation approach for accurate, fast and effective judgment and decision-making on the integrity status of train in-transit operation.

[0126] The processing flow chart of a train integrity status acquisition method based on optimized fuzzy reasoning provided by the present invention is as follows: Figure 1 As shown, the following steps are included:

[0127] Step S1, collecting the front-end satellite positioning decision information, train running speed and acceleration information of the target train through the on-board equipment at the front end of the train, collecting the rear-end satellite positioning decision information, train running speed and acceleration information of the train through the on-board equipment at the rear end of the train, and combining with the target line track space database to construct a basic feature data set of the train integrity status;

[0128] Step S2: Based on the basic feature data set of the train integrity status, the paired relative distances between the locomotives and the carriages contained in the train are collected through an ultra-wideband sensor, and the distances are associated with the train identity through time alignment to construct an enhanced feature data set of the train integrity status;

[0129] Step S3, establishing a basic fuzzy reasoning system for train integrity status decision-making based on the train integrity status enhanced feature data set, and determining a fuzzy rule table and a defuzzification rule of the basic fuzzy reasoning system;

[0130] Step S4, updating the membership function parameters of the basic fuzzy inference system based on the intelligent optimization decision logic, and establishing a train integrity status decision optimization fuzzy inference system;

[0131] Step S5: Implement integrity event alarm based on the integrity status decision result output by the optimized fuzzy inference system, and determine the integrity associated area boundary of the train based on the position of the front end and the rear end of the train.

[0132] The specific process of step S1 is:

[0133] (1) The on-board equipment at the front end of the train uses the satellite positioning terminal to collect satellite positioning decision information at the roof antenna at the front end of the train. The satellite positioning decision information includes the longitude of the roof antenna at the front end of the train at time t. latitude Elevation etc., using the included inertial measurement terminal to collect and calculate the running status information of the front end of the train, including the acceleration of the front end of the train at time t speed wait.

[0134] (2) The on-board equipment at the rear of the train uses the satellite positioning terminal contained in it to collect satellite positioning decision information from the built-in antenna at the rear end of the train. The satellite positioning decision information includes the longitude of the rear end of the train at time t. latitude Elevation etc., using the included inertial measurement terminal to collect and calculate the running status information of the rear end of the train, including the acceleration of the rear end of the train at time t speed wait.

[0135] (3) Introduce the target line track space database based on the coordinate position of the train front roof antenna Rear end of train Among them, L, B, and H represent the longitude, latitude, and elevation coordinates respectively, the superscripts H and E represent the front end and the back end respectively, and the subscript t represents the time, and the orbit frame position at the previous time is is the index (where represents the one-dimensional mileage position of the front and rear ends of the train along the track at time (t-1), and based on the search scale coefficient η, the candidate track feature point subsets of the front and rear ends of the train are retrieved and extracted forward along the train running direction. Among them, P i x Indicates the i-th track feature point at the x end within the search scale coefficient range, based on which the minimum relevant segment in the track frame corresponding to the target position (including the front and rear ends of the train) is determined Solve the front and rear track frame positions of the train at the current moment Combined with the train front roof antenna registration parameter Δ H , the internal antenna alignment parameters of the rear end of the train Δ E Calculate the estimated train length

[0136]

[0137] Figure 2 A schematic diagram of a train length estimation principle based on the front and rear positions of a train provided by an embodiment of the present invention. Figure 2 As shown in the figure, the train length directly calculated using the positioning positions of the onboard equipment at the front and rear ends of the train is the unregistered train length estimate, which has a deviation term from the actual train body length, including the uncovered part of the unregistered amount (which can be determined by the alignment parameter Δ H calibration), the over-coverage part (can be determined by the internal antenna alignment parameter Δ ETherefore, when the method described in this embodiment is used for information collection and judgment, if the above two alignment parameters are pre-measured and calibrated, the deviation term can be compensated and corrected according to formula (1), thereby obtaining a more accurate vehicle length estimation result.

[0138] (4) Based on the front track frame position of the target train Rear track frame position Front-end running speed Backend running speed Front end acceleration Back-end acceleration Estimated train length etc., to construct the basic feature dataset of train integrity status at time t

[0139] The specific process of step S2 is:

[0140] (1) The locomotives and carriages inside the train are equipped with ultra-wideband sensors to collect the relative distances between pairs of carriages in real time. Figure 3 A schematic diagram of the relative distance measurement principle between a locomotive / carriage and between carriages / carriages provided by an embodiment of the present invention is shown in FIG. Figure 3 As shown in the figure, according to the distribution of train locomotives and carriages, it is assumed that the train contains N carriages in total, the carriage closest to the locomotive is numbered CAR(1), and the carriages are numbered successively to CAR(N) in the direction away from the locomotive. There are two cases for implementing ranging information observation:

[0141] 1) Train locomotive, consisting of locomotives (for multiple locomotives, only the last locomotive is considered, Figure 3 The UWB sensor installed at the rear edge measures the distance D between the train locomotive and the adjacent carriage connected to it. LOC,t , where the measured pairwise relative distances include the offset δ between the locomotive ultra-wideband sensor and the rear edge of the locomotive loc , Carriage CAR (1) ultra-wideband sensor offset from the front edge of the car Distance between locomotive and car (1) d LOC-CAR(1),t ,Right now

[0142]

[0143] For the locomotive, only one ultra-wideband sensor is installed in the direction of the adjacent car CAR (1), and it is only necessary to pre-calibrate δ before applying this method. loc For the carriage CAR (1), an ultra-wideband sensor is installed in the direction of the adjacent locomotive and the adjacent carriage CAR (2). LOC-CAR(1),tThe relevant constraint boundary threshold judgment and reasoning analysis need to be pre-calibrated Calibration parameters in the direction of the adjacent carriage CAR(2) The calibration will be used for D LOC,t The relevant constraint boundary threshold judgment and reasoning analysis

[0144] 2) For train carriage CAR(i), where i < N, the distance D between carriage CAR(i) and its adjacent carriage CAR(i + 1) is measured by an ultra-wideband sensor installed at the rear edge of the carriage CAR(i),t , where the measured pairwise relative distance includes the offset of the ultra-wideband sensor at the rear end of carriage CAR(i) from the rear edge of the carriage The offset of the ultra-wideband sensor at the front end of carriage CAR(i + 1) from the front edge of the carriage The distance d between carriage CAR(i) and CAR(i + 1) CAR(i)-CAR(i+1),t , that is

[0145]

[0146] Among them, for carriage CAR(i), one ultra-wideband sensor is installed in the direction of the adjacent carriage CAR(i - 1) and the adjacent carriage CAR(i + 1) respectively, and pre-calibration is required Among them The calibration will be used for D CAR(i-1),t The relevant constraint boundary threshold judgment and reasoning analysis (for i = 1, it is used for d LOC-CAR(1),t ), The calibration will be used for D CAR(i),t The relevant constraint boundary threshold judgment and reasoning analysis

[0147] 3) For train carriage CAR(N), the distance D between carriage CAR(N - 1) and CAR(N) is measured by an ultra-wideband sensor installed at the rear edge of carriage CAR(N - 1) CAR(N-1),t , for carriage CAR(N), one ultra-wideband sensor is installed in the direction of the adjacent carriage CAR(N - 1), and pre-calibration is required Used for D CAR(N-1),t The relevant constraint boundary threshold judgment and reasoning analysis

[0148] (2) After the train integrity detection management platform obtains the pairwise relative distance information measured by the wireless sensors sent by the locomotive and carriages, it performs time alignment and train identity association on the relative distance information through time indexing and identity indexing respectively, where

[0149] Time alignment means that through the timestamp information in the relative distance information set, the information in the same moment and aligned with the train integrity status decision processing cycle T det is grouped into clusters, with t = tp , t=t p +T det For example, the resulting cluster is

[0150]

[0151] Train identity association refers to the secondary segmentation of the clusters aligned by time using the train ID number as the index to form an ultra-wideband measurement relative distance data set indexed by a specific time and a specific train identity. p , t=t p +T det And the train ID number is n j For example, the relative distance data set obtained by ultra-wideband ranging information observation is

[0152]

[0153] Among them, N(n j ) indicates that the ID number is n j The total number of carriages in a train excluding the locomotive.

[0154] (3) Using time information and train ID as index, The corresponding ID is n j The basic characteristic data set of train integrity status Ψ is calculated by train collection t (n j ) can be combined to form a train integrity status enhanced feature dataset. The data information items it contains can be further expanded into

[0155]

[0156] For each train in transit, by continuously collecting and extracting the enhanced feature data set of the train integrity status at each moment associated with it, real-time analysis and decision-making of the train integrity status can be further carried out, and corresponding disposal operations can be implemented based on the decision results.

[0157] The specific process of step S3 is:

[0158] (1) Based on the various information items contained in the enhanced feature dataset of train integrity status, a basic fuzzy inference system for train integrity status decision-making is established. The system is used to make a decision on the train integrity status at a specific time t and a specific train n. j To implement integrity status determination, the system mainly consists of three components:

[0159] Module 1: Information input processing module

[0160] The various types of information contained in the train integrity status enhanced feature data set are grouped into four information items to construct standardized information input rules, mainly including:

[0161] 1) Train length information item, which is estimated by the train length express.

[0162] 2) Train front and rear end speed deviation information item, which is the train front end speed collected and calculated by the train front end onboard equipment and the train rear end onboard equipment Train rear end speed Make the difference, that is

[0163]

[0164] 3) Train front and rear end acceleration deviation information item, which is the front end acceleration of the train collected by the front end on-board equipment and the rear end on-board equipment of the train Rear end acceleration Make the difference, that is

[0165]

[0166] 4) Paired relative distance information between carriages, the relative distance D obtained by using ultra-wideband ranging between locomotive / carriage and carriage / carriage γ,t (n j ) set, where γ = {LOC-CAR(1), CAR(1)-CAR(2), …, CAR(N-1)-CAR(N)}, which contains a total of N relative distance quantities.

[0167] When establishing the basic fuzzy inference system for train integrity status decision-making, considering that there will be differences in the range equivalent of train length, train speed and acceleration, and paired relative distance, which may cause difficulties in fuzzification processing and thus affect the performance and response time of the fuzzy inference system, the difference constraint with the reference value is adopted, and the information items are uniformly converted with the difference as the basic attribute, and the difference information items are used as the fuzzification input of the basic fuzzy inference system. On the one hand, the difference is easier to be fuzzified than the original value, and on the other hand, it can more effectively reflect the changes in the train integrity status. Therefore, the above four information items are parameterized by difference and constrained by range, and the corresponding parameter restriction rules are as follows:

[0168]

[0169]

[0170] in, is the nominal length of the train, ΔL, ΔV, ΔA, and ΔD are the constraints of the train length deviation, the speed deviation of the front and rear ends of the train, the acceleration deviation of the front and rear ends of the train, and the relative distance between the pairs of carriages, respectively.

[0171] Module 2: Fuzzy reasoning operation module

[0172] Based on the information item after standardized transformation as input, the triangular membership function is used as the membership function of the basic fuzzy inference system for train integrity status decision, which is used to describe the degree of belonging of each input information item element to the fuzzy set. The expression of the triangular membership function corresponding to the three cases of normal (SN), possibly normal (SM), and abnormal (SU) can be written as:

[0173]

[0174] Among them, f1 * (x), Respectively D γ,t The membership function of the four information items, {p k}, k = 1 to 4 are the right end points of the four SN triangle membership functions, {a k ,b k}, k = 1 to 4 are the left starting point, vertex, and right ending point of the four SM triangle membership functions, {q k}, k=1~4 are the left starting points of the four SU triangular membership functions, and F is the horizontal coordinate boundary value of the membership function.

[0175] Based on the setting of the above membership function, specific fuzzy rules are further established to map fuzzy input to fuzzy output. Fuzzy rules consist of a series of conditional statements, where each condition contains one or more fuzzy inputs and one fuzzy output. Fuzzy rules are determined by a specific set of fuzzy rule tables in the form of a set. For example, when the train length error is normal, the train speed error is normal, the train acceleration error is normal, and the distance between every two adjacent carriages is normal, it can be inferred that the integrity of the train is complete, and its fuzzy language expression is:

[0176]

[0177] Based on this, through the steps of fuzzification, fuzzy reasoning and defuzzification, the input information is transformed into a fuzzy description of the output, thus realizing fuzzy reasoning.

[0178] Module 3: Status determination output module

[0179] Based on the fuzzy description of the output obtained by fuzzy reasoning operation, the fuzzy output is processed through defuzzification operation in the state judgment output module to obtain a final and accurate integrity state decision result.

[0180] (2) In the process of establishing the basic fuzzy reasoning system for train integrity status decision-making, a four-input single-output system was designed. Based on the definition of fuzzy subsets, each fuzzy input has three conditions: normal (SN), possibly normal (SM), and abnormal (SU), and the fuzzy output has two conditions: complete (TN) and lost integrity (TU). The corresponding fuzzy rule table is established, which contains a total of 81 fuzzy rules, as shown in Table 1.

[0181] Table 1 Fuzzy rules of basic fuzzy inference system for train integrity status decision

[0182]

[0183]

[0184] (3) In the process of establishing the basic fuzzy reasoning system for train integrity status decision-making, the defuzzification rules adopted by the status judgment output module to implement the defuzzification operation mainly consider the activated Fuzzy rules are merged, and the output of each rule Aggregated into a single fuzzy set of output variables, the maximum operator is used to obtain

[0185]

[0186] The output of the fuzzy set calculated by the aggregation method is

[0187]

[0188] The above principles are used to finally determine the train integrity status as intact (TN) or lost integrity (TU).

[0189] The specific process of step S4 is:

[0190] (1) The selection of membership function parameters in the basic fuzzy inference system for train integrity status decision-making is optimized to improve the performance and robustness of the fuzzy inference system. The intelligent optimization decision logic is used to implement the iterative optimization of the membership function parameters, thereby establishing an optimized fuzzy inference system for train integrity status decision-making. Figure 4 A processing flow chart of updating membership function parameters based on intelligent optimization decision logic provided by an embodiment of the present invention. Figure 4 As shown, the execution of the intelligent optimization decision logic adopted by the present invention specifically includes the following steps:

[0191] Step 4.1: Initialize the population to be optimized

[0192] The population to be optimized consists of a particle set In the initialization process, the boundary of each particle should be reasonably set according to the actual situation, so as to facilitate the exploration of particles in a reasonable range and speed up the search for the optimal solution.

[0193] Step 4.2: Particle fitness evaluation

[0194] For each particle in the population, the corresponding fuzzy inference system file is generated according to the membership function parameters it represents. The acquired train integrity status enhanced feature data set is tested on the generated fuzzy inference system, and the obtained accuracy value is used as the fitness measure of each particle for feedback. The fitness can be expressed as

[0195]

[0196] Among them, TP represents true positive examples, that is, the number of positive samples correctly predicted by the system as positive; TN represents true negative examples, that is, the number of negative samples correctly predicted by the system as negative; FP represents false positive examples, that is, the number of negative samples incorrectly predicted by the system as positive; FN represents false negative examples, that is, the number of positive samples incorrectly predicted by the system as negative.

[0197] Step 4.3: Individual and global optimal updates

[0198] For each particle, first compare its current fitness with its individual historical optimal fitness. If the current fitness is better, update the individual optimal solution of the particle to further explore the potential optimal solution space. Secondly, for the entire particle set, search for the particle with the best fitness and take its fitness value as the global optimal solution. Finally, by updating the global optimal solution, guide the entire population to move towards a better solution, thereby accelerating the convergence of optimization and searching for the global optimal solution.

[0199] Step 4.4: Particle velocity and position update

[0200] According to the individual optimal solution of each particle and the global optimal solution, the particle evolution update speed and position are updated. The speed update takes into account the particle's current speed, the direction from the current position to the individual optimal position, and the direction from the current position to the global optimal position.

[0201] In order to enhance the global search capability, the fusion Lévy flight strategy is introduced to optimize the particle position update mechanism. On the basis of conventional particle position update based on current speed, individual historical optimal position and group historical optimal position, the particle position update mechanism is adjusted by introducing Lévy flight, allowing particles to explore in a long-step manner, and the step length follows the Lévy distribution, where Lévy can be expressed as:

[0202]

[0203] in, is the position of particle i in dimension j at the next iteration step (k+1), is the position of particle i in dimension j at the current iteration step k, α is the step size factor, λ is the scaling parameter of the Levy distribution, β is the shape parameter of the Levy distribution, and Γ is the Gamma function.

[0204] Step 4.5: Iteration termination condition detection

[0205] Check whether the population update iteration reaches the iteration termination condition, which includes two types:

[0206] (1) The current population iteration number has reached the maximum number of iterations I max ;

[0207] (2) The global optimal solution obtained in the current iteration step is better than the preset threshold W max .

[0208] When any one of the above two conditions is met, it indicates that the optimization iteration of the membership function parameters required to construct the train integrity status decision-making optimization fuzzy inference system has been completed, and the subsequent steps will be entered to complete the update and construction of the train integrity status decision-making optimization fuzzy inference system; when neither of the above two conditions is met, return to step 4.2 and continue the next round of iteration.

[0209] Step 4.6: Confirm the optimization parameters of the membership function

[0210] Whether the value of the membership parameter corresponding to the global optimal solution of the current iteration step meets the range condition, if the range constraint condition is met, the group of global optimal solutions is used as the output to confirm the optimal set of membership function parameters required for the train integrity status decision optimization fuzzy inference system.

[0211] (2) The optimal set of membership function parameters obtained by iterative optimization of the intelligent optimization decision logic is used to update the basic fuzzy reasoning system for train integrity status decision. The triangular membership functions corresponding to the three situations of normal (SN), possibly normal (SM), and abnormal (SU) in the fuzzy reasoning operation module are optimized. The fuzzy reasoning is implemented using the fuzzy rule table shown in Table 1, and the fuzzy output is processed through defuzzification operation in the state judgment output module to obtain a final and accurate integrity status decision result.

[0212] Figure 5 A schematic diagram for comparing train integrity status decision results provided by an embodiment of the present invention, Figure 5A set of comparison results of integrity detection decision-making based on the measured data collected by the front and rear satellite positioning receivers, inertial measurement units, and UWB sensors in the train integrity detection experiment are given. Among them, the integrity status decision results of implementing a single-channel threshold detection strategy for three types of information items, namely, train length deviation, train front and rear end running state deviation (speed deviation, acceleration deviation), and pairwise relative distance between carriages, are given for comparison. The results show that compared with the decision-making judgment based on a single category of information items, the proposed integrity status decision scheme based on fuzzy reasoning has higher sensitivity and achieves rapid detection and response 1 to 3 seconds in advance when generating early alarms; in terms of decision accuracy, compared with the single threshold strategy using only train length estimation, front and rear end state comparison, and ultra-wideband ranging, the fuzzy reasoning scheme has improved by 8.97%, 12.81%, and 2.56%, respectively. The results verify that the established optimized fuzzy reasoning method can comprehensively utilize multi-channel differentiated sensor observation information to ensure the sensitivity and accuracy of train integrity status decision-making.

[0213] Through the above steps, the established train integrity status decision optimization fuzzy inference system can automatically adjust and optimize the parameters of the membership function in the fuzzy inference system, while ensuring the performance of the fuzzy inference system and improving the system reliability and efficiency.

[0214] The specific process of step S5 is:

[0215] (1) The real-time data collected during train operation is used to update the enhanced feature data set of the train integrity status. The train integrity status decision optimization fuzzy inference system implements reasoning and decision-making. Based on the integrity status decision results output by the system, the following response operations are performed in the event of an integrity event:

[0216] 1) When the train integrity status decision optimization fuzzy inference system outputs the ID at time t, it should be n j When the integrity status of the train is complete (TN), train n j The integrity event flag remains F TIM =0;

[0217] 2) When the train integrity status decision optimization fuzzy inference system outputs the ID at time t, it should be n j When the integrity status of the train is lost integrity (TU), train n j The integrity event flag F TIM Set to 1 and lock F TIM Value modification rights, based on F TIM =1 This condition generates and outputs integrity event alarm information, which includes:

[0218] $TIMAlert, timestamp t, train ID number n j, integrity event flag F TIM =1, the first position of the alarm list

[0219] (2) When the train integrity status decision optimization fuzzy inference system implements reasoning and makes decisions and determines the integrity status output, the corresponding integrity-related area boundary is generated synchronously to assist in confirming the train's track occupancy when the train integrity status is in the state of complete (TN), and to assist in confirming the integrity-related area boundary range that needs to be protected when the front and rear ends of the train are separated when the train integrity status is in the state of losing integrity (TU). The specific principles are as follows:

[0220] 1) When the train integrity status decision optimization fuzzy inference system outputs the ID at time t, it should be n j When the integrity status of the train is complete (TN), the front and rear ends of the integrity association area boundary are calculated based on the estimated solution of the front position of the train and the length of the train. According to the availability conditions of satellite positioning at the front and rear of the train, it is determined in two cases:

[0221] Case 1: When satellite positioning is available

[0222] Integrity Association Region Boundary Front End for

[0223]

[0224] in, The track frame position calculated for the train front onboard equipment, It is the front-end margin, and its sign is determined by the up and down directions of the train.

[0225] Integrity Association Region Boundary Backend for

[0226]

[0227] in, The calculated track frame position for the rear-end onboard equipment, It is the rear-end boundary margin, and its sign is determined by the up and down directions of the train.

[0228] Case 2: When satellite positioning is not available

[0229] Integrity Association Region Boundary Front End for

[0230]

[0231] in, The train front position information is extracted from the speed and distance measurement module contained in the train control on-board equipment.

[0232] For The corresponding front-end margin, the positive or negative sign is determined by the up and down directions of the train.

[0233] Integrity Association Region Boundary Backend for

[0234]

[0235] 2) When the train integrity status decision optimization fuzzy inference system outputs the ID at time t, it should be n j When the integrity status of the train is loss of integrity (TU), the front and rear ends of the integrity association area boundary are calculated based on the estimated solution of the front and rear ends of the train under the most unfavorable conditions. According to the availability conditions of satellite positioning at the front and rear ends of the train, it is determined in two cases:

[0236] Case 1: When satellite positioning is available

[0237] Integrity Association Region Boundary Front End for

[0238]

[0239] Integrity Association Region Boundary Backend for

[0240]

[0241] Case 2: When satellite positioning is not available

[0242] Integrity Association Region Boundary Front End for

[0243]

[0244] Integrity Association Region Boundary Backend for

[0245]

[0246] in, For the train at t TU0 From the moment n TU0 Function of the distance traveled from the disconnection of the first carriage to the disconnection of the rear end of the train at the current time t, m i is the mass of the i-th carriage, χ m The most unfavorable slope condition for the target line.

[0247] Based on the method for obtaining the train integrity status based on optimized fuzzy inference provided in this embodiment, this embodiment further provides a train integrity status decision-making system based on optimized fuzzy inference, and its structure is as shown in Figure 6 Figure 1. The train integrity status decision-making system based on optimized fuzzy inference includes on-vehicle equipment at the front end of the train, on-vehicle equipment at the rear end of the train, a train unit ranging unit, and a train integrity detection management platform. The specific details are as follows:

[0248] The on-vehicle equipment at the front end of the train is composed of a satellite positioning terminal, an accelerometer, a wireless transmission module, and an on-vehicle interface unit. It is used to collect real-time satellite positioning decision-making information, train running speed, and acceleration information at the front end of the target train during operation, send the collected information to the train integrity detection management platform to construct an enhanced feature dataset for the train integrity status, and establish an optimized fuzzy inference system for train integrity status decision-making; receive integrity event warning information through the included wireless transmission module, and send the integrity event warning information to the driver's operation platform through the on-vehicle interface unit.

[0249] The on-vehicle equipment at the rear end of the train is composed of a combined antenna unit (including a satellite navigation antenna and a wireless communication antenna), a satellite positioning terminal, an accelerometer, and a wireless transmission module. It is used to collect real-time satellite positioning decision-making information, train running speed, and acceleration information at the rear end of the target train during operation, send the collected information to the train integrity detection management platform to construct an enhanced feature dataset for the train integrity status, and establish an optimized fuzzy inference system for train integrity status decision-making.

[0250] The train unit ranging unit is composed of an ultra-wideband sensor and a wireless transmission module. It is used to collect real-time paired relative distance information between the locomotive and the car body / car body and car body during operation, send the collected information to the train integrity detection management platform to construct an enhanced feature dataset for the train integrity status, and establish an optimized fuzzy inference system for train integrity status decision-making; according to the distribution composition of the train locomotive and car bodies, assuming that the train contains N car bodies in total, the deployment of the train unit monitoring unit is divided into three cases:

[0251] (1) For the train locomotive, one train unit ranging unit is installed at the rear edge of the locomotive (only considering the last locomotive in the case of multi-locomotive traction) to measure the relative distance between the locomotive and car body CAR(1).

[0252] (2) For train car body CAR(i), where i < N, one train unit ranging unit is installed at the front and rear edges of the car body respectively. Among them, the train unit ranging unit installed at the rear end of the car body measures the relative distance between this car body CAR(i) and the adjacent rear car body CAR(i + 1).

[0253] (3) Carriage CAR(N) of the train is the last car and has a train set distance measuring unit installed at the front edge of the car to measure the relative distance between CAR(N-1) and CAR(N) using the train set distance measuring unit installed at the rear end of the car in front of it, CAR(N-1).

[0254] The train integrity detection management platform is composed of a track space database server, a train integrity status analysis unit, a data recording unit, and a wireless transmission unit. It is used to receive, store, and manage observation information sent by the front-end on-board equipment, the rear-end on-board equipment, and the train group ranging unit contained in each train under its jurisdiction, build and update the basic feature data set of the train integrity status and the enhanced feature data set of the train integrity status, establish and maintain the basic fuzzy reasoning system for train integrity status decision-making and the optimized fuzzy reasoning system for train integrity status decision-making, implement train integrity status decision-making for each train under its jurisdiction, send integrity event alarms when integrity events occur, and calculate and determine the boundaries of integrity-related areas to support the implementation of corresponding integrity status response disposal.

[0255] In summary, the embodiments of the present invention fully consider the availability issues of the existing technical solutions when facing complex operating environment conditions, give full play to the ability of superposition of multi-source information conditions, integrate a variety of characteristic information related to the integrity status of the train with significant sources and characteristics, and comprehensively use the optimized fuzzy reasoning decision-making mechanism to form a new solution for train integrity status monitoring and decision-making. From the results shown in the embodiments, compared with the use of a single information channel to implement train integrity detection, better integrity status detection performance can be obtained, the response speed to the occurrence of integrity events is faster, and it can be more conducive to generating alarms and driving disposal operations when integrity events occur.

[0256] Those skilled in the art can understand that the accompanying drawings are only schematic diagrams of an embodiment, and the modules or processes in the accompanying drawings are not necessarily required to implement the present invention.

[0257] It can be known from the description of the above implementation methods that those skilled in the art can clearly understand that the present invention can be implemented by means of software plus a necessary general hardware platform. Based on such an understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which can be stored in a storage medium such as ROM / RAM, a magnetic disk, an optical disk, etc., and includes a number of instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in the various embodiments of the present invention or certain parts of the embodiments.

[0258] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device or system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment. The device and system embodiments described above are merely schematic, 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.

[0259] The above is only a preferred specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed by the present invention should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A train integrity status acquisition method based on optimized fuzzy reasoning, characterized in that: include: The front-end satellite positioning decision information, train running speed and acceleration information of the target train are collected through the on-board equipment at the front end of the train, and the rear-end satellite positioning decision information, train running speed and acceleration information are collected through the on-board equipment at the rear end of the train. Combined with the track space database of the target line, the basic characteristic data set of the train integrity status is constructed; Collecting the pairwise relative distances between the locomotives and the carriages contained in the train through an ultra-wideband sensor, associating with the train identity through time alignment, and constructing a train integrity status enhanced feature data set based on the train integrity status basic feature data set according to the pairwise relative distances between the locomotives and the carriages contained in the train; Establishing a basic fuzzy reasoning system for train integrity status decision-making based on the train integrity status enhanced feature data set, and determining a fuzzy rule table and a defuzzification rule of the basic fuzzy reasoning system; Based on the intelligent optimization decision logic, the membership function parameters of the basic fuzzy inference system are updated to establish an optimized fuzzy inference system for train integrity status decision; Implementing integrity event alarm based on the integrity status decision result output by the optimized fuzzy inference system, and determining the integrity associated area boundary of the train based on the position of the front end and the rear end of the train; The basic fuzzy inference system for train integrity status decision-making is established based on the train integrity status enhanced feature data set, and the fuzzy rule table and defuzzification rules of the basic fuzzy inference system are determined, including: Based on the various information items contained in the enhanced feature data set of train integrity status, a basic fuzzy inference system for train integrity status decision-making is established. The basic fuzzy inference system for train integrity status decision-making is used to determine the specific time t and specific train n. j Implement integrity status determination; In the process of establishing the basic fuzzy inference system for train integrity status decision, a four-input single-output system is designed. Based on the definition of fuzzy subsets, each fuzzy input has three cases: normal SN, possibly normal SM and abnormal SU, and the fuzzy output has two cases: complete TN and lost integrity TU. The corresponding fuzzy rule table is established, which contains 81 fuzzy rules in total. In the process of establishing the basic fuzzy inference system for train integrity status decision, the state judgment output module implements the defuzzification operation using the defuzzification rules to activate Fuzzy rules are merged, and the output of each rule A single set of fuzzy rules aggregated into the output variable is obtained using the maximum operator notation: The output of the fuzzy rule set calculated using the aggregation method is: Based on the membership function, a specific fuzzy rule is established, and the fuzzy rule input is mapped to the fuzzy output. Through the steps of fuzzification, fuzzy reasoning and defuzzification, the input information is converted into a fuzzy description of the output to realize fuzzy reasoning. Based on the fuzzy description of the output obtained by the fuzzy reasoning operation, the fuzzy output is processed by the defuzzification operation in the state judgment output module to obtain the train integrity state decision result as a complete TN state or a lost integrity TU state. The establishment process of the basic fuzzy inference system for train integrity status decision-making includes: The various types of information contained in the train integrity status enhanced feature data set are grouped into four information items: train length information item Train front and rear speed deviation information items Train front and rear end acceleration deviation information items and the pairwise relative distance information item D between carriages γ,t (n j ), the four information items are parameterized by difference and their ranges are constrained. The corresponding parameter restriction rules are as follows: in, is the nominal length of the train, ΔL, ΔV, ΔA, and ΔD are the constraints of the train length deviation, the speed deviation of the front and rear ends of the train, the acceleration deviation of the front and rear ends of the train, and the relative distance between the pairs of carriages, respectively; Based on the information item after standardized transformation as input, the triangular membership function is used as the membership function of the basic fuzzy inference system for train integrity status decision, which is used to describe the degree of belonging of each input information item element to the fuzzy set. The expression of the triangular membership function corresponding to the three cases of normal SN, possible normal SM and abnormal SU is: in, Respectively D γ,t The membership function of the four information items, {p k }, k = 1 to 4 are the right end points of the four SN triangle membership functions, {a k ,b k ,c k }, k = 1 to 4 are the left starting point, vertex, and right ending point of the four SM triangle membership functions, {q k }, k=1~4 are the left starting points of the four SU triangular membership functions, and F is the horizontal coordinate boundary value of the membership function.

2. The method according to claim 1, characterized in that: The front-end satellite positioning decision information, train running speed and acceleration information of the target train are collected by the on-board equipment at the front end of the train, and the rear-end satellite positioning decision information, train running speed and acceleration information are collected by the on-board equipment at the rear end of the train, and the basic characteristic data set of the train integrity status is constructed in combination with the track space database of the target line, including: The on-board equipment at the front end of the train uses the satellite positioning terminal contained in it to collect satellite positioning decision information at the roof antenna at the front end of the train. The satellite positioning decision information includes the longitude of the roof antenna at the front end of the train at time t. latitude and elevation Use the inertial measurement terminal to collect and calculate the running status information of the front end of the train, which includes the acceleration of the front end of the train at time t. and speed The on-board equipment at the rear of the train uses the satellite positioning terminal contained in it to collect satellite positioning decision information at the built-in antenna at the rear end of the train. The satellite positioning decision information includes the longitude of the rear end of the train at time t. latitude and elevation Use the inertial measurement terminal to collect and calculate the running status information of the rear end of the train, which includes the acceleration of the rear end of the train at time t. and speed Based on the coordinate position of the antenna on the roof of the front end of the train Rear end of train Track frame position at the previous moment Based on the search scale coefficient η, the candidate track feature point subsets of the front and rear ends of the train are retrieved and extracted forward along the train running direction. Among them, P i x Indicates the i-th track feature point at the x end within the search scale coefficient range, based on which the minimum relevant segment in the track frame corresponding to the target position is determined Solve the front and rear track frame positions of the train at the current moment Combined with the train front roof antenna registration parameter Δ H , the internal antenna alignment parameters of the rear end of the train Δ E Calculate the estimated train length: (4) Construct the basic characteristic data set of the train integrity status at time t based on the front and rear track frame positions, running speed, acceleration and train length estimation information of the target train 3. The method according to claim 2, characterized in that The method of collecting the paired relative distances between the locomotives and the carriages contained in the train through the ultra-wideband sensor, associating with the train identity through time alignment, and constructing the train integrity status enhanced feature data set based on the train integrity status basic feature data set according to the paired relative distances between the locomotives and the carriages contained in the train, includes: Ultra-wideband sensors are installed in the locomotives and carriages inside the train to collect the relative distances between pairs of carriages in real time. According to the distribution of locomotives and carriages, it is assumed that the train contains N carriages. The nearest locomotive carriage is numbered CAR(1), and the carriages are numbered successively to CAR(N) in the direction away from the locomotive. There are two cases for distance measurement information observation: 1) Train locomotive, the distance D between the train locomotive and the adjacent carriage connected to it is measured by the ultra-wideband sensor installed on the rear edge of the locomotive LOC,t , where the measured pairwise relative distances include the distance d between the locomotive and the carriage CAR(1) LOC-CAR(1),t ,Right now: For the locomotive, an ultra-wideband sensor is installed in the direction of the adjacent car CAR (1), and the pre-calibrated δ loc For the carriage CAR (1), an ultra-wideband sensor is installed in the direction of the adjacent locomotive and the adjacent carriage CAR (2). LOC-CAR(1),t Related constraint threshold judgment and reasoning analysis, pre-calibration Calibration parameters in the direction of the adjacent car CAR(2) The calibration will be used for D LOC,t Related constraint threshold judgment and reasoning analysis; 2) The train carriage CAR(i), where i < N, the distance D between the carriage CAR(i) and its adjacent carriage CAR(i + 1) is measured by an ultra-wideband sensor installed at the rear edge of the carriage CAR(i),t , where the measured pairwise relative distances include the offset of the ultra-wideband sensor at the rear end of the carriage CAR(i) from the rear edge of the carriage the offset of the ultra-wideband sensor at the front end of the carriage CAR(i + 1) from the front edge of the carriage the distance d between the carriages CAR(i) and CAR(i + 1) CAR(i)-CAR(i+1),t , that is For car CAR(i), an ultra-wideband sensor is installed in the direction of the adjacent car CAR(i-1) and the adjacent car CAR(i+1), and the pre-calibrated in, The calibration will be used for D CAR(i-1),t Related constraint threshold judgment and reasoning analysis, The calibration will be used for D CAR(i),t Related constraint threshold judgment and reasoning analysis; 3) Train car CAR(N), the distance D between car CAR(N-1) and CAR(N) is measured by the ultra-wideband sensor installed on the rear edge of car CAR(N-1) CAR(N-1),t For the car CAR(N), an ultra-wideband sensor is installed in the direction of the adjacent car CAR(N-1) and pre-calibrated For D CAR(N-1),t Related constraint threshold judgment and reasoning analysis; (2) After the train integrity detection management platform obtains the paired relative distance information measured by the wireless sensors sent by the locomotive and the carriage, it aligns the relative distance information with the train identity through the time index and identity index respectively, where: Time alignment refers to the decision processing cycle T of the train integrity status that is aligned at the same time by using the timestamp information in the relative distance information set det The information is divided into groups and clustered, with t=t p , t=t p +T det For example, the resulting cluster is in, t p Time information cluster, Indicates that the ID is n j The distance between the locomotive of a train and the adjacent carriage connected to it is t p The measured distance at a moment, Indicates that the ID is n k The distance between the locomotive of a train and the adjacent carriage connected to it is t p The measured distance at a moment, Indicates that the ID is n x The distance between the i-th car of the train and the adjacent (i+1)-th car connected to it is t p The measured distance at a moment, Indicates that the ID is n y The distance between the u-th car of the train and the adjacent (u+1)-th car connected to it is t p The measured distance at a moment, Train identity association refers to the secondary segmentation of the clusters aligned by time using the train ID number as the index to form an ultra-wideband measurement relative distance data set indexed by a specific time and a specific train identity. p , t=t p +T det And the train ID number is n j For example, the relative distance data set obtained by observing ultra-wideband ranging information is: in, Indicates t p The time is based on the train ID number n j is the relative distance dataset obtained by indexing, It means (t p +T det ) time with train ID number n j is the relative distance dataset obtained by index, N(n j ) indicates that the ID number is n j The total number of carriages in a train excluding locomotives, Indicates that the ID is n j The distance between the locomotive of a train and the adjacent carriage connected to it is t p The measured distance at a moment, Indicates that the ID is n j The distance between the first carriage of the train and the second adjacent carriage connected to it is t p The measured distance at a moment, Indicates that the ID is n j The train number (N(n j )-1) car and the adjacent N(n)th car connected to it j ) Between the carriages at t p The measured distance at a moment, Indicates that the ID is n j The distance between the locomotive of a train and the adjacent carriage connected to it is (t p +T det ) at the time of measurement, Indicates that the ID is n j The distance between the first carriage of the train and the second adjacent carriage connected to it is (t p +T det ) at the time of measurement, Indicates that the ID is n j The train number (N(n j )-1) car and the adjacent N(n)th car connected to it j ) carriages between (t p +T det ) at the moment of measurement; (3) Using time information and train ID as index, The corresponding ID is n j The basic characteristic data set of train integrity status Ψ is calculated by train collection t (n j ) to form a train integrity status enhanced feature dataset The data information items it contains are further expanded as follows: Among them, D loc,t (n j ) indicates that the ID is n j The measured distance between the locomotive of a train and the adjacent carriage connected to it at time t, D CAR(1),t (n j ) indicates that the ID is n j The measured distance between the first carriage of the train and the second adjacent carriage connected to it at time t, Indicates that the ID is n j The train number (N(n j )-1) car and the adjacent N(n)th car connected to it j ) is the measured distance between the carriages at time t, Indicates that the ID is n j The front track frame position of the train at time t, Indicates that the ID is n j The front running speed of the train at time t, Indicates that the ID is n j The front acceleration of the train at time t is Indicates that the ID is n j The rear track frame position of the train at time t, Indicates that the ID is n j The rear running speed of the train at time t, Indicates that the ID is n j The rear end acceleration of the train at time t is Indicates that the ID is n j The estimated length of the train at time t.

4. The method according to claim 1, characterized in that The method of updating the membership function parameters of the basic fuzzy inference system based on the intelligent optimization decision logic to establish an optimized fuzzy inference system for train integrity status decision-making includes: The intelligent optimization decision logic is used to implement the iterative optimization determination of the membership function parameters of the basic fuzzy inference system. The execution of the intelligent optimization decision logic specifically includes the following steps: Step 4.1: Initialize the population to be optimized The population to be optimized consists of a particle set Composition, wherein each particle represents a set of triangular membership function parameter sets, including control point parameters of each fuzzy set membership function; Step 4.2: Particle fitness evaluation For each particle in the population, the corresponding fuzzy inference system file is generated according to the membership function parameters represented by the particle. The acquired train integrity status enhanced feature data set is tested on the generated fuzzy inference system, and the obtained accuracy value is used as the fitness measure of each particle for feedback. The fitness is expressed as: Among them, TP represents true positive examples, that is, the number of positive samples correctly predicted by the system as positive; TN represents true negative examples, that is, the number of negative samples correctly predicted by the system as negative; FP represents false positive examples, that is, the number of negative samples incorrectly predicted by the system as positive; FN represents false negative examples, that is, the number of positive samples incorrectly predicted by the system as negative; Step 4.3: Individual and global optimal updates For each particle, first compare its current fitness with its individual historical optimal fitness. If the current fitness is better, update the individual optimal solution of the particle to further explore the potential optimal solution space. Secondly, for the entire particle set, search for the particle with the best fitness and take its fitness value as the global optimal solution. Finally, by updating the global optimal solution, guide the entire population to move towards a better solution. Step 4.4: Particle velocity and position update According to the individual optimal solution of each particle and the global optimal solution, the particle evolution update speed and position are updated. The speed update takes into account the particle's current speed, the direction from the current position to the individual optimal position, and the direction from the current position to the global optimal position; Step 4.5: Iteration termination condition detection Check whether the population update iteration reaches the iteration termination condition, which includes two types: (1) The current population iteration number has reached the maximum number of iterations I max ; (2) The global optimal solution obtained in the current iteration step is better than the preset threshold W max ; When any one of the above two conditions is met, it indicates that the optimization iteration of the membership function parameters required to construct the train integrity status decision optimization fuzzy inference system has been completed, and the subsequent steps will be entered to complete the update and construction of the train integrity status decision optimization fuzzy inference system; when both of the above two conditions are not met, return to step 4.2 and continue the next round of iteration; Step 4.6: Confirm the optimization parameters of the membership function Whether the value of the membership parameter corresponding to the global optimal solution of the current iteration step meets the range condition, if the range constraint condition is met, the group of global optimal solutions is used as the output to confirm the optimal set of membership function parameters required for the train integrity status decision optimization fuzzy inference system; The optimal set of membership function parameters obtained by iterative optimization of intelligent optimization decision logic is used to update the basic fuzzy reasoning system for train integrity status decision, and the triangular membership functions corresponding to the three situations of normal (SN), possibly normal (SM), and abnormal (SU) in the fuzzy reasoning operation module are optimized. The fuzzy reasoning is implemented using the fuzzy rule table shown in Table 1, and the fuzzy output is processed through defuzzification operation in the state judgment output module to obtain a final and accurate integrity status decision result.

5. The method according to claim 4, characterized in that The method of implementing integrity event alarm based on the integrity status decision result output by the optimized fuzzy inference system, and determining the integrity associated area boundary of the train based on the position of the front end and the rear end of the train, includes: (1) The real-time data collected during train operation is used to update the enhanced feature data set of the train integrity status. The optimized fuzzy inference system for train integrity status decision-making implements reasoning and decision-making. Based on the integrity status decision results output by the system, the following response operations are performed in the event of an integrity event: 1) When the train integrity status decision optimization fuzzy inference system outputs the ID at time t, it should be n j When the integrity status of the train is complete TN, train n j The integrity event flag remains F TIM =0; 2) When the train integrity status decision optimization fuzzy inference system outputs the ID at time t, it should be n j When the integrity state of the train is lost integrity TU, train n j The integrity event flag F TIM Set to 1 and lock F TIM Value modification rights, based on F TIM =1 This condition generates integrity event alarm information and outputs it; (2) When the optimized fuzzy inference system of the train integrity status decision implements the reasoning decision and determines the integrity status output, the corresponding integrity association area boundary is generated synchronously to assist in confirming the train's track occupancy when the train integrity status is in the complete TN state, and to assist in confirming the integrity association area boundary range that needs to be protected when the front and rear ends of the train are separated when the train integrity status is in the lost integrity TU state. The specific principles are as follows: 1) When the train integrity status decision optimization fuzzy inference system outputs the ID at time t, it should be n j When the integrity status of the train is complete TN, the front and rear ends of the integrity association area boundary are calculated based on the estimated solution of the front position of the train and the length of the train. According to the availability conditions of the satellite positioning of the front and rear ends of the train, it is determined in two cases: Case 1: When satellite positioning is available Integrity Association Region Boundary Front End for: in, The track frame position calculated for the train front onboard equipment, is the front-end margin, and its sign is determined by the up and down directions of the train. Integrity Association Region Boundary Backend for: in, The calculated track frame position for the rear-end onboard equipment, is the rear-end margin, and its sign is determined by the up and down directions of the train. Case 2: When satellite positioning is not available Integrity Association Region Boundary Front End for: in, The train front position information is extracted from the speed and distance measurement module contained in the train control on-board equipment. For The corresponding front-end margin, the sign is determined by the up and down direction of the train; Integrity Association Region Boundary Backend for: 2) When the train integrity status decision optimization fuzzy inference system outputs the ID at time t, it should be n j When the integrity status of the train is loss of integrity (TU), the front and rear ends of the integrity association area boundary are calculated based on the estimated solution of the front and rear ends of the train under the most unfavorable conditions. According to the availability conditions of the satellite positioning of the front and rear ends of the train, they are determined in two cases: Case 1: When satellite positioning is available Integrity Association Region Boundary Front End for: Integrity Association Region Boundary Backend for: Case 2: When satellite positioning is not available Integrity Association Region Boundary Front End for: Integrity Association Region Boundary Backend for: in, For the train at t TU0 From the moment n TU0 Function of the distance traveled from the disconnection of the first carriage to the disconnection of the rear end of the train at the current time t, m i is the mass of the i-th carriage, χ m The most unfavorable slope condition for the target line.

Images