A heavy-haul train group cooperative control method, device, equipment and medium

Abstract

The present disclosure relates to the field of rail transit technology, and particularly relates to a heavy-haul train group cooperative control method and device, equipment and medium. The present disclosure determines the safety of the heavy-haul train cooperative control environment according to the segmented data of the laser radar and the speed and distance of the detected object to the solid-state laser radar. When the solid-state laser radar detects that the cooperative control environment is safe, that is, within the cooperative control distance range, in addition to the heavy-haul train in front, there is no other object, the solid-state laser radar will obtain the actual distance and speed of the heavy-haul train in front for cooperative control, and use the distance error and speed of the heavy-haul train as input and the speed regulation gain of the heavy-haul train as output to establish a heavy-haul train cooperative control regulation system, and optimize the minimum error adaptive fuzzy PID controller based on the improved particle swarm optimization algorithm, thereby shortening the distance of the heavy-haul train cooperative control, improving the railway transportation capacity, and effectively solving the shortcomings of the traditional control system in the dynamic environment.

Description

Technical Field

[0001] This disclosure relates to the field of rail transit technology, and in particular to a method, device, equipment and medium for coordinated control of heavy-load train groups. Background Technology

[0002] With the rapid development of railway transportation, especially heavy-haul trains, the cooperative control technology of train groups has become crucial to ensuring efficient and safe transportation. Traditional train control systems face challenges in handling cooperative control of train groups, including environmental perception accuracy, data processing efficiency, control algorithm stability, and adaptability to dynamic environments, making it difficult to maintain precise train spacing and speed regulation in complex environments. Currently, accurate distance measurement has become a key factor in virtual coupled maneuvering and vehicle-to-vehicle (V2V) systems.

[0003] Existing technologies do not fully consider the dynamic conditions, train composition, speed and distance changes required for virtual coupled trains, especially in maneuvering and V2V coupled operations in complex environments. Existing systems struggle to handle precise train spacing and dynamic adaptability. Summary of the Invention

[0004] To address the aforementioned issues, this disclosure provides a method, apparatus, equipment, and medium for coordinated control of heavy-haul train groups, which shortens the coordinated control distance of heavy-haul trains and improves railway transport capacity.

[0005] A first aspect includes a method for coordinated control of heavy-haul train groups, the method comprising:

[0006] The following train in the heavy-haul train group uses lidar to acquire real-time segmented data of the surrounding environment of the heavy-haul train group;

[0007] The real-time segmented data is preprocessed using dynamic programming to obtain preprocessed data.

[0008] A fuzzy clustering algorithm based on dynamic thresholds is used to fuzz the preprocessed data, extract the objects of interest, and obtain the distances between the following vehicle and all objects of interest within a preset distance range.

[0009] When there is only a heavy-load train ahead within the preset distance range, an adaptive fuzzy PID controller that minimizes error is used to adjust the distance and speed between the following train and the heavy-load train ahead, based on the distance between the following train and the heavy-load train ahead.

[0010] Furthermore, real-time segmented data includes:

[0011] The relative distance between the object of interest and the heavy-duty train equipped with lidar, the speed of the object of interest relative to the heavy-duty train equipped with lidar, and the segment distance of the solid-state lidar.

[0012] Among them, the relative distance between the object of interest and the heavy-haul train equipped with lidar and the speed of the object of interest relative to the heavy-haul train equipped with lidar are directly measured;

[0013] The segment distance D of the solid-state lidar is calculated using the following formula:

[0014]

[0015] Where D represents the segment distance of the lidar, d i and d i+1 Let θ and θ represent the distance between the two laser beams emitted by the solid-state lidar, respectively; let cos(θ) represent the cosine value of angle θ; and let i represent the number of segments of the lidar.

[0016] Furthermore, dynamic programming is used to preprocess the real-time segmented data, including:

[0017] The collected lidar segmented data is preprocessed using dynamic programming to generate the corresponding binary membership matrix as preprocessed data.

[0018] The formula for dynamic programming is as follows:

[0019]

[0020] Where D = [D1, D2, ..., D i [] represents the distance data between segments of a solid-state lidar, where i represents the number of lidar segments, and D... i Let represent the distance data of the i-th segment of the LiDAR, diff[i] represent the difference between the i-th and m-th adjacent distance data points, diff[m] represent the difference between the m-th and m-th adjacent distance data points, j represent the starting index of the LiDAR distance data segment, dp[i] represent the optimal grouping cost of the first i segments, dp[j] represent the optimal grouping cost of the first j segments, and cost(j+1,i) represent the cost from segment D. j+1 To D i The cost and, j * Let j represent the optimal cut point with the minimum total cost, P represent the set of all optimal cut points, and j1 represent the optimal cut point of the first segment group. S-1 Let S represent the optimal cut point of the (S-1)th segment group, where S represents the optimal number of segment groups. Let j represent the o-th optimal cutting point. o+1 The value of |P| represents the starting position of segment o+1; |P| represents the absolute value of P; argmin represents the value of the variable that minimizes the expression, i.e., the value of the optimal segment point; the formula for the binary membership matrix is ​​as follows:

[0021]

[0022] Among them, M p,q This represents the q-th data point in the p-th segment, i.e., an element in the data matrix, where p indicates the current segment is p, and q represents the q-th element in the current data point sequence. q B represents an element in the data sequence, specifically the q-th data point in the input data. p,q This represents the q-th binary affiliation matrix element in the p-th segment.

[0023] Furthermore, the adaptive fuzzy clustering algorithm based on dynamic thresholds is formulated as follows:

[0024]

[0025] Where σ represents the σ-th iteration, l represents the l-th column of the pregrouping matrix (i.e., the l-th segment), and r represents the r-th row of the pregrouping matrix (i.e., the r-th segment group). This indicates the center position of the r-th segment group in the σ-th iteration. D represents the degree of belonging to the l-th segment in the r-th segment group after the σ-th iteration pre-grouping. l x represents the segment measurement distance of the current l-th LiDAR segment. lr Let represent the distance between LiDAR segments in the l-th segment of the r-th segment group of the pre-grouped matrix, where i represents the number of LiDAR segments, and S represents the optimal number of segment groups. Let v represent the degree of belonging to the l-th segment of the r-th segment group of the σ-th iteration binary matrix, and let v represent the fuzzy index, Δc. r This represents the change in the center of the r-th segment group after t iterations, where ò0 represents the initial threshold. σ The convergence threshold representing the dynamic change in the σ-th iteration, error. σ The error represents the error or change value in the σ-th iteration. σ α is used to measure the state of the algorithm in the σth iteration, where α represents the adjustment factor.

[0026] Furthermore, a fuzzy clustering algorithm based on dynamic thresholds is used to perform fuzzy processing on the preprocessed data to extract objects of interest, including:

[0027] A fuzzy clustering algorithm based on dynamic thresholds is used to perform fuzzy processing on the preprocessed data to obtain a fuzzy membership matrix that represents the membership relationship between objects and laser beams.

[0028] The number of objects within the field of view of the lidar is determined by the fuzzy membership matrix, as well as the lidar segment group corresponding to the object. Then, the distance to the object is obtained according to the lidar ranging formula.

[0029] The cluster closest to the lidar sensor is selected as the object of interest.

[0030] Furthermore, adjustments are made to the distance and speed between the following train and the heavy-haul train ahead, including:

[0031] The distance error value is calculated based on the distance between the following train and the preceding heavy-haul train, and the expected cooperative distance. The speed of the heavy-haul train is then adjusted using cooperative control based on this distance error value. The cooperative control adjustment is expressed as follows:

[0032]

[0033] Among them, D lateral,min d represents the three-dimensional distance of the object closest to the lidar. ref L represents the expected collaborative distance. delay L represents the distance traveled by the following train when there is a communication delay with adjacent trains. break(r) L represents the maximum braking distance of the following vehicle. break(p) L represents the maximum braking distance of the vehicle in front. s The value represents the safety margin, max() represents taking the maximum value, t represents time, e(t) represents the error between the actual distance of the train and the target distance at time t, and d actual (t) represents the actual coordinated distance of the train at time t. T represents the rate of change of error e(t). sample Represents the sampling period, e(tT) sample ) represents the error at the previous sampling time, v(t) represents the speed of the train at time t, v ref (t) represents the expected cooperative velocity, v ref (t) is obtained from the driving diagram, where K represents the gain coefficient, which controls the speed's response to distance error, and S... k S represents the distance from idle stop. e The common braking distance of the train operation monitoring and recording device LKJ is represented by v0, the initial braking speed is represented by λ, and the braking calculation coefficient is represented by ν. h Indicates the train switching braking rate, φ h β represents the friction coefficient converted from brake shoe. c This represents common braking parameters, where ω0 represents the train's basic unit resistance, t0 represents the driver's reaction time, and v... m This indicates the final speed of a heavily loaded train after it has slowed down.

[0034] Furthermore, the adaptive fuzzy PID controller that minimizes the error is formulated as follows:

[0035]

[0036] Where u(t) represents the output signal of the controller at time t, e(t) represents the error between the actual distance of the train and the target distance at time t, and t represents time. K represents the rate of change of error e(t). υ K represents the proportional gain parameter. β K represents the integral gain parameter. d K represents the differential gain parameter. υ (t) represents the proportional gain at time t, K d (t) represents the differential gain at time t, K β (t) represents the integral gain at time t, K υ0 K represents the initial proportional gain. β0 K represents the initial integral gain. d0 Let ΔK represent the initial differential gain. υ ΔK represents the proportional gain adjustment amount. d ΔK represents the differential gain adjustment. β ΔK represents the integral gain adjustment amount. υ ΔK d ΔK β The three adjustment values ​​were obtained through fuzzy reasoning.

[0037] Furthermore, it also includes:

[0038] An improved particle swarm optimization algorithm is used to dynamically adjust the initial parameters of the adaptive fuzzy PID controller with minimum error.

[0039] The particle update strategy of the improved particle swarm optimization algorithm is as follows:

[0040]

[0041] Among them, H n K represents the position of the nth particle in the current search space. υ K represents the proportional gain parameter. β K represents the integral gain parameter. d Let represent the differential gain parameter, T represent the time step range for error calculation, Q(z) represent the error at iteration z, J represent the objective function to be minimized, and v n (z) represents the velocity of the nth particle in the z-th iteration, w represents the inertia weight term, and p n Let r1 and r2 represent random numbers between [0,1], c1 represent the first learning factor, c2 represent the second learning factor, g represent the global optimal position found among all particles, and x represent the current optimal position found among all particles. n (z) The position of particle n at iteration z, w max and w minThese represent the maximum and minimum values ​​of the inertia weight, respectively, and z represents the number of iterations. max Maximum number of iterations These represent the maximum values ​​of c1 and c2, respectively. Let c1 and c2 represent the minimum values, respectively.

[0042] Secondly, a collaborative control device for heavy-haul train groups includes: a segmented data acquisition unit, a preprocessing unit, an object of interest extraction unit, and a collaborative control unit;

[0043] The segmented data acquisition unit is used to acquire real-time segmented data of the surrounding environment of the heavy-haul train group based on lidar in the following train in the heavy-haul train group.

[0044] The preprocessing unit is used to preprocess real-time segmented data using dynamic programming to obtain preprocessed data.

[0045] The object of interest extraction unit is used to perform fuzzy processing on the preprocessed data using a fuzzy clustering algorithm based on dynamic thresholds, extract the object of interest, and obtain the distance between the following vehicle and all objects of interest within a preset distance range;

[0046] The cooperative control unit is used to adjust the distance and speed between the following vehicle and the heavy-load train ahead based on the distance between the following vehicle and the heavy-load train ahead, using an adaptive fuzzy PID controller that minimizes errors, when there is only a heavy-load train ahead within a preset distance range.

[0047] Furthermore, real-time segmented data includes:

[0048] The distance from the object of interest to the lidar, the speed of the object of interest relative to the heavy-duty train carrying the lidar, and the segment distance of the solid-state lidar;

[0049] Among them, the distance from the object of interest to the lidar and the speed of the object of interest relative to the heavy-duty train equipped with the lidar are directly measured by the solid-state lidar;

[0050] The segmented distance D of the solid-state lidar is calculated using the following formula:

[0051]

[0052] Where D represents the segment distance of the lidar, d i and d i+1 Let θ and θ represent the distance between the two laser beams emitted by the solid-state lidar, respectively; let cos(θ) represent the cosine value of angle θ; and let i represent the number of segments of the lidar.

[0053] Furthermore, the preprocessing unit is specifically used for:

[0054] The collected lidar segmented data is preprocessed using dynamic programming to generate the corresponding binary membership matrix as preprocessed data.

[0055] The formula for dynamic programming is as follows:

[0056]

[0057] Where D = [D1, D2, ..., D i [] represents the distance data between segments of a solid-state lidar, where i represents the number of lidar segments, and D... i Let represent the distance data of the i-th segment of the LiDAR, diff[i] represent the difference between the i-th and m-th adjacent distance data points, diff[m] represent the difference between the m-th and m-th adjacent distance data points, j represent the starting index of the LiDAR distance data segment, dp[i] represent the optimal grouping cost of the first i segments, dp[j] represent the optimal grouping cost of the first j segments, and cost(j+1,i) represent the cost from segment D. j+1 To D i The cost and, j * Let j represent the optimal cut point with the minimum total cost, P represent the set of all optimal cut points, and j1 represent the optimal cut point of the first segment group. S-1 Let S represent the optimal cut point of the (S-1)th segment group, where S represents the optimal number of segment groups. Let j represent the o-th optimal cutting point. o+1 The value of |P| represents the starting position of segment o+1; |P| represents the absolute value of P; argmin represents the value of the variable that minimizes the expression, i.e., the value of the optimal segment point; the formula for the binary membership matrix is ​​as follows:

[0058]

[0059] Among them, M p,q This represents the q-th data point in the p-th segment, i.e., an element in the data matrix, where p indicates the current segment is p, and q represents the q-th element in the current data point sequence. q B represents an element in the data sequence, specifically the q-th data point in the input data. p,q This represents the q-th binary affiliation matrix element in the p-th segment.

[0060] Furthermore, the adaptive fuzzy clustering algorithm based on dynamic thresholds is formulated as follows:

[0061]

[0062] Where σ represents the σ-th iteration, l represents the l-th column of the pregrouping matrix (i.e., the l-th segment), and r represents the r-th row of the pregrouping matrix (i.e., the r-th segment group). This indicates the center position of the r-th segment group in the σ-th iteration. D represents the degree of belonging to the l-th segment in the r-th segment group after the σ-th iteration pre-grouping. l x represents the segment measurement distance of the current l-th LiDAR segment. lr Let represent the distance between LiDAR segments in the l-th segment of the r-th segment group of the pre-grouped matrix, where i represents the number of LiDAR segments, and S represents the optimal number of segment groups. Let v represent the degree of belonging to the l-th segment of the r-th segment group of the σ-th iteration binary matrix, and let v represent the fuzzy index, Δc. r This represents the change in the center of the r-th segment group after t iterations, where ò0 represents the initial threshold. σ The convergence threshold representing the dynamic change in the σ-th iteration, error. σ The error represents the error or change value in the σ-th iteration. σ α is used to measure the state of the algorithm in the σth iteration, where α represents the adjustment factor.

[0063] Furthermore, the object of interest extraction unit is specifically used for:

[0064] A fuzzy clustering algorithm based on dynamic thresholds is used to perform fuzzy processing on the preprocessed data to obtain a fuzzy membership matrix that represents the membership relationship between objects and laser beams.

[0065] The number of objects within the field of view of the lidar is determined by the fuzzy membership matrix, as well as the lidar segment group corresponding to the object. Then, the distance to the object is obtained according to the lidar ranging formula.

[0066] The cluster closest to the lidar sensor is selected as the object of interest.

[0067] Furthermore, the collaborative control unit is specifically used for:

[0068] The distance error value is calculated based on the distance between the following train and the preceding heavy-haul train, and the expected cooperative distance. The speed of the heavy-haul train is then adjusted using cooperative control based on this distance error value. The cooperative control adjustment is expressed as follows:

[0069]

[0070] Among them, D lateral,min d represents the three-dimensional distance of the object closest to the lidar. ref L represents the expected collaborative distance. delay L represents the distance traveled by the following train when there is a communication delay with adjacent trains. break(r)L represents the maximum braking distance of the following vehicle. break(p) L represents the maximum braking distance of the vehicle in front. s The value represents the safety margin, max() represents taking the maximum value, t represents time, e(t) represents the error between the actual distance of the train and the target distance at time t, and d actual (t) represents the actual coordinated distance of the train at time t. T represents the rate of change of error e(t). sample Represents the sampling period, e(tT) sample ) represents the error at the previous sampling time, v(t) represents the speed of the train at time t, v ref (t) represents the expected cooperative velocity, v ref (t) is obtained from the driving diagram, where K represents the gain coefficient, which controls the speed's response to distance error, and S... k S represents the distance from idle stop. e The common braking distance of the train operation monitoring and recording device LKJ is represented by v0, the initial braking speed is represented by λ, and the braking calculation coefficient is represented by ν. h Indicates the train switching braking rate, φ h β represents the friction coefficient converted from brake shoe. c This represents common braking parameters, where ω0 represents the train's basic unit resistance, t0 represents the driver's reaction time, and v... m This indicates the final speed of a heavily loaded train after it has slowed down.

[0071] Furthermore, the adaptive fuzzy PID controller that minimizes the error is formulated as follows:

[0072]

[0073] Where u(t) represents the output signal of the controller at time t, e(t) represents the error between the actual distance of the train and the target distance at time t, and t represents time. K represents the rate of change of error e(t). υ K represents the proportional gain parameter. β K represents the integral gain parameter. d K represents the differential gain parameter. υ (t) represents the proportional gain at time t, K d (t) represents the differential gain at time t, K β (t) represents the integral gain at time t, K υ0 K represents the initial proportional gain. β0 K represents the initial integral gain. d0 Let ΔK represent the initial differential gain. υ ΔK represents the proportional gain adjustment amount. d ΔK represents the differential gain adjustment.β ΔK represents the integral gain adjustment amount. υ ΔK d ΔK β The three adjustment values ​​were obtained through fuzzy reasoning.

[0074] Furthermore, an improved particle swarm optimization algorithm is used to dynamically adjust the initial parameters of the adaptive fuzzy PID controller with minimum error.

[0075] The particle update strategy of the improved particle swarm optimization algorithm is as follows:

[0076]

[0077] Among them, H n K represents the position of the nth particle in the current search space. υ K represents the proportional gain parameter. β K represents the integral gain parameter. d Let represent the differential gain parameter, T represent the time step range for error calculation, Q(z) represent the error at iteration z, J represent the objective function to be minimized, and v n (z) represents the velocity of the nth particle in the z-th iteration, w represents the inertia weight term, and p n Let r1 and r2 represent random numbers between [0,1], c1 represent the first learning factor, c2 represent the second learning factor, g represent the global optimal position found among all particles, and x represent the current optimal position found among all particles. n (z) The position of particle n at iteration z, w max and w min These represent the maximum and minimum values ​​of the inertia weight, respectively, and z represents the number of iterations. max Maximum number of iterations These represent the maximum values ​​of c1 and c2, respectively. Let c1 and c2 represent the minimum values, respectively.

[0078] Thirdly, an electronic device includes a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus;

[0079] Memory, which stores computer programs;

[0080] When a processor executes a computer program stored in memory, it implements the aforementioned method for coordinated control of heavy-load train groups.

[0081] Fourthly, a computer-readable storage medium stores a computer program that, when executed by a processor, implements the aforementioned method for coordinated control of heavy-load train groups.

[0082] This disclosure includes at least the following beneficial effects:

[0083] This disclosure achieves precise adjustment of train spacing and speed by processing data acquired by lidar in real time and combining dynamic programming and fuzzy control algorithms. Through an adaptive fuzzy PID controller and an improved particle swarm optimization algorithm, controller parameters are dynamically optimized, improving system stability and control accuracy. This effectively addresses the shortcomings of traditional control systems in dynamic environments, providing a new solution for the efficient and safe operation of heavy-haul train groups.

[0084] This disclosure determines the safety of the cooperative control environment for heavy-haul trains based on segmented data from a lidar, the speed of detected objects, and the distance to the solid-state lidar. When the solid-state lidar detects a safe cooperative control environment (i.e., within the cooperative control distance range, there are no other objects besides the heavy-haul train ahead), the solid-state lidar acquires the actual distance and speed of the heavy-haul train under cooperative control. Using the distance error and speed of the heavy-haul train as input and the speed adjustment gain of the heavy-haul train as output, a cooperative control and adjustment system for heavy-haul trains is established. Based on the improved particle swarm optimization algorithm (IPSO), the adaptive fuzzy PID controller with minimum error is optimized, shortening the cooperative control distance for heavy-haul trains and improving railway transport capacity.

[0085] Other features and advantages of this disclosure will be set forth in the following description and will be apparent in part from the description or may be learned by practicing the disclosure. The objects and other advantages of this disclosure may be realized and obtained by means of the structures pointed out in the description and the accompanying drawings. Attached Figure Description

[0086] To more clearly illustrate the technical solutions in the embodiments of this disclosure or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this disclosure. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0087] Figure 1 This is a schematic diagram of the control method flow according to an embodiment of the present disclosure;

[0088] Figure 2 This is a schematic diagram of the control device structure according to an embodiment of the present disclosure;

[0089] Figure 3 This is a schematic diagram of the electronic device structure according to an embodiment of the present disclosure;

[0090] Figure 4 This is a schematic diagram of the segmented distance of the lidar in an embodiment of this disclosure;

[0091] Figure 5 This is a convergence curve of the dynamic threshold method according to an embodiment of this disclosure;

[0092] Figure 6 This is a radar test field of view diagram according to an embodiment of the present disclosure, wherein the object is in the middle of the field of view and close to the lidar;

[0093] Figure 7 This is a radar test field of view diagram according to an embodiment of the present disclosure, wherein the object is in the middle of the field of view and far away from the lidar;

[0094] Figure 8 This is a radar test field of view diagram according to an embodiment of the present disclosure, wherein only half of the object is within the field of view of the lidar;

[0095] Figure 9 This is an embodiment of the present disclosure. Figure 6 A schematic diagram of the initial matrix corresponding to the radar test field of view;

[0096] Figure 10 This is an embodiment of the present disclosure. Figure 6 A schematic diagram of the fuzzy matrix corresponding to the radar test field of view;

[0097] Figure 11 This is an embodiment of the present disclosure. Figure 7 A schematic diagram of the initial matrix corresponding to the radar test field of view;

[0098] Figure 12 This is an embodiment of the present disclosure. Figure 7 A schematic diagram of the fuzzy matrix corresponding to the radar test field of view;

[0099] Figure 13 This is an embodiment of the present disclosure. Figure 8 A schematic diagram of the initial matrix corresponding to the radar test field of view;

[0100] Figure 14 This is an embodiment of the present disclosure. Figure 8 A schematic diagram of the fuzzy matrix corresponding to the radar test field of view;

[0101] Figure 15 This is a block diagram of the cooperative control of heavy-haul trains according to an embodiment of this disclosure;

[0102] Figure 16 This invention discloses a three-dimensional diagram showing the membership degree, speed, and cooperative distance relationship of a heavily loaded train ahead, measured by a train behind it.

[0103] Figure 17 This is a diagram showing the distance error and real-time distance error rate of the collaborative control of heavy-haul trains according to an embodiment of this disclosure;

[0104] Figure 18 This is a diagram showing the speed error and real-time speed error rate of the collaboratively controlled heavy-haul train according to an embodiment of this disclosure;

[0105] Figure 19 This is a diagram showing the cooperative speed and cooperative distance of the heavy-haul trains under cooperative control according to an embodiment of this disclosure;

[0106] Figure 20 This is a schematic diagram of the IPSO algorithm optimization process according to an embodiment of this disclosure;

[0107] Figure 21 This diagram illustrates a comparison of the iterative changes in the optimal individual fitness value between the improved IPSO algorithm and the traditional PSO algorithm, as shown in the embodiments of this disclosure. Detailed Implementation

[0108] To make the objectives, technical solutions, and advantages of the embodiments of this disclosure clearer, the technical solutions of the embodiments of this disclosure will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this disclosure, and not all embodiments. Based on the embodiments of this disclosure, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this disclosure.

[0109] like Figure 1 As shown, a method for coordinated control of heavy-haul train groups includes:

[0110] S101, the following train in the heavy-haul train group obtains real-time segmented data of the surrounding environment of the heavy-haul train group based on lidar;

[0111] S102, The real-time segmented data is preprocessed using dynamic programming to obtain preprocessed data;

[0112] S103, a fuzzy clustering algorithm based on dynamic threshold is used to perform fuzzy processing on the preprocessed data, extract the objects of interest, and obtain the distance between the following vehicle and all objects of interest within a preset distance range;

[0113] S104 When there is only a heavy-load train ahead within the preset distance range, the following car uses an adaptive fuzzy PID controller with minimized error to adjust the distance and speed between the following car and the heavy-load train ahead, based on the distance between the following car and the heavy-load train ahead.

[0114] The specific implementation details are as follows:

[0115] Based on real-time segmented data acquired by solid-state lidar, the surrounding environment of heavy-haul trains is perceived with high precision to ensure environmental safety for collaborative control.

[0116] Based on the distance error obtained from the solid-state lidar for the coordinated control of heavy-haul trains, the distance and speed of the coordinated control of heavy-haul trains are adjusted.

[0117] Dynamic programming is used to preprocess the real-time segmented data acquired from the solid-state lidar. To more accurately identify the front-to-back relationships of objects, an adaptive fuzzy clustering algorithm based on dynamic thresholds is further employed to fuzzify the preprocessed segmented data, thereby extracting the target objects of interest. The distance from the solid-state lidar to the target object of interest is calculated. An adaptive fuzzy PID controller with minimized error is used to precisely adjust the coordinated control distance and speed of the heavy-haul train. The parameters of the adaptive fuzzy PID controller with minimized error can be dynamically adjusted. An improved particle swarm optimization algorithm is used to dynamically adjust the initial parameters of the adaptive fuzzy PID controller with minimized error.

[0118] The segmented data includes: the segmented distance of the solid-state lidar, the distance from the object of interest to the lidar, and the speed of the heavy-haul train. The distance from the object of interest to the lidar and the speed of the heavy-haul train are measured directly by the solid-state lidar. The segmented distance of the lidar is expressed as:

[0119]

[0120] Where D represents the segment distance of the lidar, d i and d i+1 Let θ and θ represent the distance between the two laser beams emitted by the solid-state lidar, respectively; let cos(θ) represent the cosine value of angle θ; and let i represent the number of segments of the lidar.

[0121] Dynamic programming is used to preprocess the collected LiDAR segmented data to generate the corresponding binary membership matrix. The dynamic programming method is expressed as:

[0122]

[0123] Where D = [D1, D2, ..., D i [] represents the distance data between segments of a solid-state lidar, where i represents the number of lidar segments, and D i Let represent the distance data of the i-th segment of the LiDAR, diff[i] represent the difference between the i-th adjacent distance data points, j represent the starting index of the LiDAR distance data segment, dp[i] represent the optimal grouping cost of the first i segments, dp[j] represent the optimal grouping cost of the first j segments, and cost(j+1,i) represent the cost from segment D. j+1 To D i The cost and, j *Let j denote the optimal cut point with the minimum total cost, and P denote the set of all optimal cut points. S-1 Let S represent the optimal cut point of the (S-1)th segment group, where S represents the optimal number of segment groups. Let j represent the k-th optimal cutting point. k+1 This indicates the starting position of segment k+1.

[0124] The formula for generating the binary affiliation matrix is:

[0125]

[0126] Among them, M p,q Let D represent the q-th data point (an element in the data matrix) in the p-th segment, where p indicates the current segment is p, and q indicates the current element is the q-th element in the data point sequence. q This represents the elements D = [D1, D2, ..., D] in the data sequence. i ], that is, the q-th data point in the input data, B p,q This represents the q-th binary matrix element in the p-th segment.

[0127] An adaptive fuzzy clustering algorithm based on dynamic thresholds is used to perform fuzzy processing on the preprocessed segmented data. The adaptive fuzzy clustering algorithm based on dynamic thresholds is expressed as follows:

[0128]

[0129] Where σ represents the σ-th iteration, l represents the l-th column (l-th segment) of the pregrouping matrix, and r represents the r-th row (j-th segment group) of the pregrouping matrix. This indicates the center position of the r-th segment group in the σ-th iteration. D represents the degree of belonging to the l-th segment in the r-th segment group after the σ-th iteration pre-grouping. i x represents the segmented measurement distance of the current i-th segment of the lidar system. lr Let represent the distance between LiDAR segments in the l-th segment of the r-th segment group of the pre-grouped matrix, where i represents the number of LiDAR segments, and S represents the optimal number of segment groups. The σ-th iteration binary matrix represents the degree of belonging to the l-th segment of the r-th segment group, where v is the fuzzy exponent, usually greater than 1, and Δc is the degree of belonging to the l-th segment. r This represents the change in the center of the r-th segment group after t iterations, where ò0 represents the initial threshold. σ The convergence threshold representing the dynamic change in the σ-th iteration, error. σ The error or change value in the σ-th iteration is used to measure the state of the algorithm in the σ-th iteration, and α represents the adjustment factor.

[0130] The cluster closest to the lidar sensor is chosen as the object of interest primarily because it offers higher security guarantees, as the closest object can be considered the most potentially dangerous. The fuzzy membership matrix is ​​defined as follows:

[0131]

[0132] Among them, u (σ) Let κ represent the fuzzy attribution matrix, where κ represents the κth column of the fuzzy attribution matrix, which in practice means that there are κ objects within the field of view of the lidar detection, and r represents the rth row of the fuzzy attribution matrix, which is the rth segment group.

[0133] The number of objects within the field of view of the lidar is determined based on the fuzzy membership matrix, and the lidar segment group corresponding to each object is identified. Then, the distance to the object is obtained according to the lidar ranging formula. The lidar ranging steps are as follows:

[0134] For the κ-th object, in order to ensure data validity while avoiding the introduction of too much irrelevant or noisy data, all laser beams with a fuzzy membership degree greater than 0.6 and meeting the angle condition are selected. The formula for the effective beam set is as follows:

[0135]

[0136] Among them, M′ κ This represents the set of effective beams for the κ-th object. φ represents the membership value between the κ-th object and the r-th laser beam segment, where r represents the r-th segment group. r φ represents the elevation angle of the r-th segment group. min Minimum elevation angle of segment group, φ min Take -15°, φ max φ represents the maximum elevation angle of the segment group. max Take 15°.

[0137] Based on the effective beam set of the object, the coordinate transformation of the lidar is then performed, using the following formula:

[0138]

[0139] Where L represents the lidar coordinate system, P L,r Let d represent the coordinates of the r-th laser beam segment in the lidar coordinate system. r The distance of the r-th laser segment, θ r φ represents the horizontal angle of the r-th segment group. r Let B represent the elevation angle of the r-th segment group, and let B represent the vehicle coordinate system. The original coordinates of the r-th laser segment in the vehicle coordinate system, where R represents the rotation matrix, which can be obtained from the installation angle of the laser radar relative to the vehicle, and T represents the translation vector, i.e. the offset of the coordinate origin, which can be obtained through calibration.

[0140] After the transformation from the radar coordinate system to the vehicle coordinate system, the resulting coordinates represent the position of the segment group in the vehicle coordinate system. These beam data require further processing for more accurate fusion and target localization, as shown in the following formula:

[0141]

[0142] Where L represents the lidar coordinate system, B represents the vehicle coordinate system, and r represents the r-th segment group. The weight of the r-th beam segment, φ r φ represents the elevation angle of the r-th segment group. opt Let λ represent the optimal angle, which is 0°, λ represent the attenuation factor, and δ represent the smoothing constant. This indicates the fusion position of the κ-th object. M′ represents the initial position of the r-th laser segment. κ Let v represent the set of valid segments for the κ-th object. B,r Let v represent the velocity of the r-th laser segment in the vehicle coordinate system. r,doppler This represents the Doppler velocity of the r-th laser segment. T represents the calibrated position of the κ-th object after velocity and acceleration compensation. sample This indicates the radar sampling period.

[0143] The obtained positions are in vector form. To convert them into actual planar distances, the formula is as follows:

[0144]

[0145] Where B represents the vehicle coordinate system. This represents the calibrated position of the κ-th object after velocity and acceleration compensation. This indicates the lateral position of the object in the vehicle coordinate system with Doppler velocity compensation. This indicates the longitudinal position of the object in the vehicle coordinate system with Doppler velocity compensation. This indicates the vertical position of the object in the vehicle coordinate system with Doppler velocity compensation. This indicates the object's lateral position in the vehicle coordinate system without Doppler velocity compensation. This indicates the longitudinal position of the object in the vehicle coordinate system without Doppler velocity compensation. This indicates the vertical position of the object in the vehicle coordinate system without Doppler velocity compensation, T. sampleThis represents the sampling period, where r represents the r-th segment. The weight of the r-th segment group, M′ κ Let v represent the set of valid segments for the κ-th object. B,r,x Let v represent the lateral velocity of the r-th laser segment in the vehicle coordinate system. B,r,y Let v represent the longitudinal velocity of the r-th laser segment in the vehicle coordinate system. B,r,z Let r represent the vertical velocity of the r-th laser segment in the vehicle coordinate system.

[0146] The cluster closest to the lidar sensor is taken as the object of interest, as shown in the following formula:

[0147]

[0148] Among them, D lateral,κ The three-dimensional distance of the laser beam to the κ-th object. This indicates the lateral position of the object in the vehicle coordinate system with Doppler velocity compensation. This indicates the longitudinal position of the object in the vehicle coordinate system with Doppler velocity compensation. D represents the vertical position of the object in the vehicle coordinate system with Doppler velocity compensation. lateral,min Let M' represent the 3D distance to the object closest to the lidar, r represent the r-th segment, and M′ represent the distance to the object closest to the lidar. κ Let represent the set of effective beams for the κ-th object, and min indicates taking the minimum value.

[0149] The distance error is calculated based on the distance from the train to the solid-state lidar and the expected cooperative distance. This distance error is further calculated by subtracting the expected distance from the train's distance to the lidar. The speed of the heavy-haul train is then adjusted based on the distance error value. The cooperative control adjustment is expressed as follows:

[0150]

[0151] Among them, D lateral,min d represents the three-dimensional distance of the object closest to the lidar. ref L represents the expected collaborative distance. delay L represents the distance traveled by the following train when there is a communication delay with adjacent trains. break(r) L represents the maximum braking distance of the following vehicle. break(p) L represents the maximum braking distance of the vehicle in front. s The value represents the safety margin, max() represents taking the maximum value, t represents time, e(t) represents the error between the actual distance of the train and the target distance at time t, and d actual (t) represents the actual coordinated distance of the train at time t. T represents the rate of change of error e(t). sample Represents the sampling period, e(tT)sample ) represents the error at the previous sampling time, v(t) represents the speed of the train at time t, v ref (t) represents the expected cooperative velocity, v ref (t) is obtained from the driving diagram, where K represents the gain coefficient, which controls the speed's response to distance error, and S... k S represents the distance from idle stop. e The common braking distance of the train operation monitoring and recording device LKJ is represented by v0, the initial braking speed is represented by λ, and the braking calculation coefficient is represented by ν. h Indicates the train switching braking rate, φ h β represents the friction coefficient converted from brake shoe. c This represents common braking parameters, where ω0 represents the train's basic unit resistance, t0 represents the driver's reaction time, and v... m This indicates the final speed of a heavily loaded train after it has slowed down.

[0152] The goal of adjusting the gain coefficient K through real-time feedback is to keep the error at a minimum while avoiding over-adjustment and oscillation. The adaptive fuzzy PID controller for minimizing error continuously adjusts the gain based on the error and its rate of change. The adaptive fuzzy PID controller for minimizing error is expressed as:

[0153]

[0154] Where u(t) represents the output signal of the controller at time t, e(t) represents the error between the actual distance of the train and the target distance at time t, and t represents time. K represents the rate of change of error e(t). υ K represents the proportional gain parameter. β K represents the integral gain parameter. d K represents the differential gain parameter. υ (t) represents the proportional gain at time t, K d (t) represents the differential gain at time t, K β (t) represents the integral gain at time t, K υ0 K represents the initial proportional gain. β0 K represents the initial integral gain. d0 Let ΔK represent the initial differential gain. υ ΔK represents the proportional gain adjustment amount. d ΔK represents the differential gain adjustment. β ΔK represents the integral gain adjustment amount. υ ΔK d ΔK β The three adjustment values ​​were obtained through fuzzy reasoning.

[0155] The improved particle swarm optimization algorithm is used to dynamically adjust the initial parameters of the adaptive fuzzy PID controller to minimize error, including:

[0156] An improved particle swarm optimization algorithm is used to dynamically adjust the initial parameters of the adaptive fuzzy PID controller with minimum error. The particle update strategy of the improved particle swarm optimization algorithm is expressed as follows:

[0157]

[0158] Among them, H n K represents the position of the nth particle in the current search space. υ K represents the proportional gain parameter. β K represents the integral gain parameter. d Let represent the differential gain parameter, T represent the time step range for error calculation, Q(z) represent the error at iteration z, J represent the objective function to be minimized, and v n (z) represents the velocity of the nth particle in the z-th iteration, w represents the inertia weight term, and p n Let r1 and r2 represent random numbers between [0,1], c1 represent the first learning factor, c2 represent the second learning factor, g represent the global optimal position found among all particles, and x represent the current optimal position found among all particles. n (z) The position of particle n at iteration z, w max and w min These represent the maximum and minimum values ​​of the inertia weight, respectively, and z represents the number of iterations. max Maximum number of iterations These represent the maximum values ​​of c1 and c2, respectively. Let c1 and c2 represent the minimum values, respectively.

[0159] In the detection of the cooperative control environment of heavy-haul trains, the safety of the cooperative control environment of heavy-haul trains is determined by the segmented data of the lidar, the speed of the detected object, and the distance to the solid-state lidar. When the cooperative control environment of heavy-haul trains is safe, the data obtained by the solid-state lidar is the actual cooperative control distance and speed of the heavy-haul train in front.

[0160] In the coordinated control and regulation of heavy-haul trains, a coordinated control and regulation system for heavy-haul trains is established by taking the distance error and speed of the heavy-haul trains as inputs and the speed regulation gain of the heavy-haul trains as outputs.

[0161] In the initial parameter tuning section, an improved particle swarm optimization algorithm is used to optimize the adaptive fuzzy PID controller to minimize errors, and the optimized controller is used to control the heavy-haul train cooperative control system. The PSO optimizes three parameters in total.

[0162] like Figure 2 As shown, a collaborative control device for heavy-haul train groups includes: a segmented data acquisition unit 201, a preprocessing unit 202, an object of interest extraction unit 203, and a collaborative control unit 204.

[0163] The segmented data acquisition unit 201 is used to acquire real-time segmented data of the surrounding environment of the heavy-haul train group based on lidar in the following train in the heavy-haul train group.

[0164] Preprocessing unit 202 is used to preprocess real-time segmented data using dynamic programming to obtain preprocessed data;

[0165] The object of interest extraction unit 203 is used to perform fuzzy processing on the preprocessed data using a fuzzy clustering algorithm based on dynamic thresholds, extract the object of interest, and obtain the distance between the following vehicle and all objects of interest within a preset distance range;

[0166] The cooperative control unit 204 is used to adjust the distance and speed between the following vehicle and the heavy-load train ahead based on the distance between the following vehicle and the heavy-load train ahead when there is only a heavy-load train ahead within a preset distance range. The following vehicle uses an adaptive fuzzy PID controller with minimized error to adjust the distance and speed between the following vehicle and the heavy-load train ahead.

[0167] like Figure 3 As shown, this disclosure provides an electronic device, including a processor 301, a communication interface 302, a memory 303, and a communication bus 304, wherein the processor 301, the communication interface 302, and the memory 303 communicate with each other through the communication bus 304;

[0168] Memory 303 stores computer programs;

[0169] The processor 301 implements the above method when executing a computer program stored in the memory 303.

[0170] This disclosure provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method.

[0171] The computer-readable storage medium may be included in the device / apparatus described in the above embodiments; or it may exist independently and not assembled into the device / apparatus. The computer-readable storage medium carries one or more programs that, when executed, implement the method according to the embodiments of this disclosure.

[0172] According to embodiments of this disclosure, the computer-readable storage medium can be a non-volatile computer-readable storage medium, such as including, but not limited to: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this disclosure, the computer-readable storage medium can be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0173] To enable those skilled in the art to better understand this disclosure, the principles of this disclosure are explained below in conjunction with the accompanying drawings:

[0174] Firstly, solid-state lidar segmentation technology is introduced into the cooperative control of heavy-haul trains. The sampling time of the solid-state lidar is 10ms, and the specific parameters are shown in Table 1. The real-time segmented data acquired by the lidar is preprocessed using dynamic programming. Then, a fuzzy clustering algorithm based on dynamic thresholds is used to fuzzify the preprocessed segmented data, generating a corresponding fuzzy membership matrix. The safety of the cooperative control environment is then determined based on the fuzzy membership matrix.

[0175] Table 1

[0176] Number of segments / segments 8 Field of vision 20° Geometric dimensions 70mm×35.9mm×71.2mm Using wavelength / nm 905 Distance accuracy / cm 5 Operating temperature / °C -40～85 Maximum detection distance / m 1315

[0177] By analyzing the number of segments of the solid-state lidar (segment distance of the solid-state lidar, distance from the heavy-haul train to the solid-state lidar, distance from the obstacle to the solid-state lidar, and speed of the heavy-haul train), the feasibility of the dynamic programming method and the fuzzy clustering algorithm based on dynamic thresholds was verified, enabling accurate identification of the cooperative environment of heavy-haul trains.

[0178] In the collaborative control of heavy-haul trains, the segmented distances (such as those obtained by solid-state lidar) Figure 4 The segmented distance diagram acquired by the solid-state lidar is real-time, which requires the algorithm to process the data quickly and accurately. Dynamic programming for the segmented distance acquired by the solid-state lidar should follow these steps:

[0179] Step 1: Calculate the adjacent differences between the segmented distances acquired by the solid-state lidar. The adjacent differences are expressed as:

[0180] diff[i] = |D i+1 -D i | (15)

[0181] Where, diff[i] represents the distance difference between adjacent segments of the solid-state lidar, D i+1 and Di These represent the segment distances of the (i+1)th segment and the ith segment, respectively.

[0182] Step 2: Initialize the dynamic programming array. Initialize dp[0] = 0, which means that the grouping cost is 0 when there is no segmentation.

[0183] Step 3: Define segmented costs: Here, cost(j+1, i) is from segment D. i+1 To D i The cumulative difference.

[0184] Step 4: Dynamic programming recursion, the recursion formula is as follows:

[0185]

[0186] For each i, iterate through all possible segment starting points j and calculate the segment cost cot(j+1,i) from j+1 to i. Add the optimal cost of the current segment to the cumulative optimal cost dp[j] of the previous segments to obtain the group cost. Find the minimum replacement cost among all possible j and assign it to dp[i].

[0187] Step 5: Calculate the optimal split point, using the following formula:

[0188]

[0189] Find the cut point j that minimizes dp[j]. * And through j * Backtrack to determine the final split point.

[0190] Step 6: Calculate the optimal segment: P = {j1, j2, ..., jj} S-1 P is used to store all optimal cut points, indicating the optimal segmentation positions of the data.

[0191] Step 7: Calculate the optimal number of segments, using the following formula:

[0192] S = |P| + 1 (18)

[0193] Calculate the final number of segments, i.e., the optimal S-segment.

[0194] Step 8: Backtrack to calculate the optimal split point, using the following formula:

[0195]

[0196] Backtracking to find each optimal cutting point, determining the first segment, the second segment, and so on, up to the Sth segment.

[0197] Step 9: Output the optimal segment. After the dynamic programming is completed, the value of array dp[i] is the optimal grouping cost for the entire LiDAR segment data.

[0198] Step 10: Construct the data matrix, using the following formula:

[0199]

[0200] Fill in the corresponding value for each data segment, and fill in 0 for the rest. In the matrix M structure, the first row contains the data for the first segment, and fill in 0 for the rest. The second row contains the data for the second segment, and fill in 0 for the rest. And so on, until the Sth segment.

[0201] Step 11: Construct a binary matrix, using the following formula:

[0202]

[0203] To mark which segment a data point belongs to, matrix B has the following structure: First row: 1 for the first segment, 0 for the rest. Second row: 1 for the second segment, 0 for the rest. And so on, up to the Sth segment.

[0204] By combining the above steps, the segmented distance data acquired in real time by solid-state lidar can be processed quickly and accurately, improving the accuracy of fuzzy processing.

[0205] Fuzzy clustering algorithms based on dynamic thresholds include:

[0206] Fuzzy clustering is used to fuzzify the piecewise distances processed by dynamic programming. An accurate and stable fuzzy membership matrix is ​​obtained through iterative methods. To avoid insufficient iterations or getting stuck in infinite iteration, a dynamic threshold is used during the iteration process: First, the initial threshold is determined.

[0207] ò0=Δc r0 (twenty two)

[0208] Where ò0 represents the initial threshold, Δc r0 This represents the error value at the initial moment.

[0209] The dynamic threshold formula is:

[0210]

[0211] Where ò0 represents the initial threshold, and ò represents the convergence threshold at the start of the algorithm. σ It is a dynamically changing convergence threshold, error. σ The error or change value of the current iteration is used to measure the state of the algorithm in the σ-th iteration, α. σThis represents the dynamic adjustment coefficient, α0 represents the initial adjustment coefficient (usually 1 or other positive value), and error. max This represents the maximum value of the error, usually taken as the initial error value. σ The error value of the current iteration.

[0212] According to the dynamic threshold formula:

[0213] when error σ When the error is large, α σ Larger, making the threshold ò σ The adjustment range is larger, when error σ When the error is small, α σ Reduce, make the threshold σ The adjustment range is precise, when error σ Error less than dynamic threshold σ If the algorithm is considered to have converged, the loop is stopped. To prevent this, a maximum number of iterations σ is set. max When σ > σ max Force the loop to stop.

[0214] According to the dynamic threshold formula:

[0215] when error σ When the error is large, α σ Larger, making the threshold ò σ The adjustment range is larger, when error σ When the error is small, α σ Reduce, make the threshold σ The adjustment range is precise, when error σ Error less than dynamic threshold σ If the algorithm is considered to have converged, the loop is stopped. To prevent this, a maximum number of iterations σ is set. max When σ > σ max Force the loop to stop.

[0216] like Figure 5 As shown, the dynamic threshold method converges faster.

[0217] like Figures 6-8 The lidar field of view diagrams for the three test scenarios respectively; Figure 6 The object is in the center of the field of view, close to the lidar. Figure 7 The object is in the center of the field of view, far from the lidar. Figure 8 The object is only half within the lidar's field of view and is close to the lidar. To better simulate real-world situations, three scenarios were simulated to verify the algorithm's recognition accuracy.

[0218] Figure 9-10 The initial matrix and the fuzzy matrix correspond to Figure 6 In this case, the membership matrix consists of three clusters. Figure 9 For the initial matrix, Figure 10 The membership matrix is ​​the result of fuzzification. Because the objects are close to the lidar and all objects are within the field of view, the membership values ​​do not change significantly compared to the initial matrix. This means that the intervals between clusters are large enough that the membership values ​​in the boundary measurements are not fuzzified.

[0219] Figure 11-12 The initial matrix and the fuzzy matrix correspond to Figure 7 The situation, Figure 11 For the initial matrix, Figure 12 The blurred attribution matrix shows a slightly different situation in the second test. In this case, there is still an object in the middle, but closer to the background wall.

[0220] Figure 13-14 The initial matrix and the fuzzy matrix correspond to Figure 8 In this situation, Figure 13 For the initial matrix, Figure 14 The assigned distance matrix after fuzzification is shown. This part of the object is introduced into the LiDAR field of view, and a common phenomenon can be observed: when the object only covers a portion of the field of view, the measured distance of segment 5 is usually between the distances of the first and last clusters. In this case, the clustering algorithm will typically assign this segment to a specific cluster. Due to the characteristics of fuzzy clustering, the assignment of this intermediate value can be flexibly adjusted between two adjacent clusters, making the measurement results closer to the actual situation.

[0221] Figure 15 The block diagram of the heavy-haul train cooperative control according to an embodiment of this disclosure is shown below:

[0222] Step 103: The adaptive fuzzy PID controller with minimized error is used to precisely adjust the cooperative control distance and speed of the heavy-haul train. When the cooperative distance does not reach the expected cooperative distance, the adaptive fuzzy PID controller with minimized error will adjust the speed adjustment gain to change the speed of the heavy-haul train.

[0223] The accumulated experience and knowledge of drivers are modeled using a fuzzy rule base to establish an offline query fuzzy matrix table. During online operation, an adaptive fuzzy PID controller with minimized error is designed using fuzzy inference. Through result processing, table lookup, and computation of fuzzy logic rules, the impact of uncertainties in the coordinated control of heavy-load trains during train operation is reduced. The input and output membership functions of the fuzzy controller are both overlapping and symmetric generalized bell-shaped membership functions. The design of the adaptive fuzzy PID controller with minimized error is based on the coordinated distance error e(t) of heavy-load trains and the rate of change of the coordinated distance error. For input, ΔK p ΔK i ΔK d The change in parameters is the output. The inference uses the maximum-minimum composition rule, and the output uses the centroid defuzzification method, as detailed below:

[0224]

[0225] Where, μ(Δk') j ), Δk' j R represents the membership degree of the corresponding fuzzy subset membership function and the fuzzy inference output. j For fuzzy relationships, Δk adaptively adjusts the increment, s j This represents the membership degree in a fuzzy rule.

[0226] The adaptive fuzzy PID controller with minimized error utilizes fuzzy control principles to adjust PID parameters online in real time, enabling the controlled object to exhibit good adaptability and control performance. A suitable fuzzy rule base should be established, and its fuzzy control rules should follow these principles:

[0227] To better perform fuzzy control, thresholds are set. When a certain threshold is reached, corresponding adjustments are made. Threshold settings include:

[0228] Error range definition: Error range e range It is the cooperative distance d ref ±2%, the error range is expressed as:

[0229] e range =±0.02·d ref (25)

[0230] Where e range Indicates the error range, d ref This indicates the set cooperative distance.

[0231] The error rate of change is defined as follows:

[0232]

[0233] in, e represents the rate of change of error. range Indicates the error range, T sample Indicates the sampling period, d ref This indicates the set cooperative distance.

[0234] The error range and the error rate of change range are combined with the input design of fuzzy control and divided into three fuzzy sets: "small" (S), "medium" (M), and "large" (L), as detailed below:

[0235] Error thresholds are divided into 3 levels:

[0236] Small (S): |e|≤0.5·e range This indicates that the error is close to the target value, and the controller does not need to be significantly adjusted.

[0237] Medium (M): 0.5 e range <|e|≤0.8·e range This indicates that the error deviates from the target value by a moderate range, and the controller needs to be adjusted accordingly.

[0238] Large (L): |e|>0.8·e range This indicates that the error deviates significantly from the target value, and the controller needs to be adjusted quickly.

[0239] The error rate of change is divided into three levels:

[0240] Little (S): This indicates that the rate of error change is slow, and the controller only needs minor adjustments.

[0241] Middle (M): This indicates that the rate of error change is moderate, and the controller needs to be adjusted appropriately.

[0242] Large (L): This indicates that the error changes drastically, and the controller needs to respond quickly.

[0243] Based on the error e and the rate of change of error, a fuzzy rule is designed to adjust the PID parameters, and ΔK is designed. p ΔK i ΔK d The fuzzy rule tables are shown in Tables 2, 3, and 4:

[0244] Table 2

[0245]

[0246] Table 3

[0247]

[0248]

[0249] Table 4

[0250]

[0251] The fuzzy rule design features dynamic adaptability and robustness, and can adjust the error e and the rate of change of error in real time. Dynamically adjust PID parameters (K) p ,K i ,K dThis approach achieves a balance between rapid response and stability. It simplifies the control of complex systems through rule tables, eliminating the need for precise modeling, and offers excellent scalability and interpretability. It performs exceptionally well in handling dynamic and uncertain scenarios, such as the cooperative control of heavy-haul trains.

[0252] like Figure 16 As shown, two heavy-haul trains are operating in coordination at 50 km / h. The train behind measures the membership degree and speed of the heavy-haul train in front. From the three-dimensional diagram of the coordination distance relationship, we can see that the coordination distance is 461m and the coordination speed is 50 km / h.

[0253] like Figures 17-18 As shown, the adaptive fuzzy PID controller with minimized error performs well in the coordinated control of heavy-haul trains, with an average speed error rate of 1.18% and an average spacing error rate of 1.17%.

[0254] like Figure 19 As shown, the relationship between the cooperative distance and cooperative speed of heavy-haul trains based on lidar cooperative control can be seen, which can greatly reduce the cooperative distance and improve the railway's transport capacity.

[0255] Step 104: The initial parameters of the adaptive fuzzy PID controller with minimum error are dynamically adjusted using an improved particle swarm optimization algorithm.

[0256] This embodiment employs an adaptive fuzzy PID controller design that minimizes error and uses an improved PSO (Programmable Filtering Oscillator) to dynamically adjust the controller's initial parameters. The improved PSO for dynamically adjusting the controller's initial parameters includes:

[0257] like Figure 20 The flowchart shown is for the controller optimization based on the improved IPSO algorithm:

[0258] The procedure for setting parameters in the improved IPSO algorithm is as follows:

[0259] ① Initialization: Set the particle population size to SwarmSize = 30, and the inertia weight w max =0.9, w min =0.2, learning factor c 1min =0.1, c 2min =0.2, c 1max =1,c 2max =2, maximum search space V max =1, minimum value V min =-1, the maximum number of iterations is MaxIter=20, and the position and velocity of the particles are initialized at the same time.

[0260] ② To obtain the optimal solution, the time integral performance index of the absolute error is used as the fitness function of the IPSO algorithm, with the following criteria:

[0261]

[0262] In the formula, F is the fitness value, t is time, and e(t) is the systematic error.

[0263] ③ Generate a particle swarm, and the particle swarm affects K. p ,K i ,K d The fitness values ​​of each particle are assigned according to formula (14), and then compared and selected to find the individual best and the global best.

[0264] ④ Update the position and velocity of the particles. If the termination condition is met, output the optimal solution; otherwise, perform a particle update operation to generate a new particle swarm for the next cycle.

[0265] The variable universe fuzzy PID control parameters K are obtained by optimizing the particle swarm optimization algorithm before and after the improvement. p ,K i ,K d ,Depend on Figure 21 The results of the fitness value optimization comparison show that the improved particle swarm optimization algorithm achieves the optimal fitness value after 9 iterations. This indicates that the improved particle swarm optimization algorithm proposed in this disclosure has better optimization performance, and the optimization results better meet the optimization objective requirements.

[0266] Although the present disclosure has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present disclosure.

Claims

1. A method for coordinated control of heavy-haul train groups, characterized in that, The method includes: The following train in the heavy-haul train group uses lidar to acquire real-time segmented data of the surrounding environment of the heavy-haul train group; The real-time segmented data is preprocessed using dynamic programming to obtain preprocessed data. A fuzzy clustering algorithm based on dynamic thresholds is used to fuzz the preprocessed data, extract the objects of interest, and obtain the distances between the following vehicle and all objects of interest within a preset distance range. When there is only a heavy-load train ahead within the preset distance range, an adaptive fuzzy PID controller with minimized error is used to adjust the distance and speed between the following car and the heavy-load train ahead, based on the distance between the following car and the heavy-load train ahead. The real-time segmented data is preprocessed using dynamic programming, including: The collected lidar segmented data is preprocessed using dynamic programming to generate the corresponding binary membership matrix as preprocessed data. The formula for dynamic programming is as follows: in, This represents the distance data between segments of a solid-state lidar system. This indicates the number of segments in the lidar system. The first laser radar Segmented distance data, Indicates the first The difference between the m-th and m-th adjacent data points, where diff[m] represents the difference between the m-th and m-th adjacent data points. This represents the starting index of the lidar range data segment, where dp[i] indicates the previous segment. The optimal grouping cost for segmentation, dp[j] represents the cost of the first segment. The optimal grouping cost for segmentation, cost(j+1,i) represents the cost from segmentation. arrive The cost and, This represents the optimal split point that minimizes the total cost. Let j represent the set of all optimal cut points, and j1 represent the optimal cut point of the first segment group. Indicates the first Optimal cutting point of segment group This represents the optimal number of segments. Indicates the first The optimal cutting point express The starting position of the segment; | | represents the absolute value of P; This represents the value of the variable that minimizes the expression, i.e., the value of the optimal segmentation point; the formula for the binary membership matrix is ​​as follows: in, Indicates the first The first in the segment Each data point is an element in the data matrix. This indicates that the current number is the [number]. Segmentation, Indicates the first data point in the current data point sequence. One element, This represents an element in the data sequence, specifically the first element in the input data. Data points, Indicates the first The first in the segment Each binary affiliation matrix element; A fuzzy clustering algorithm based on dynamic thresholds is used to perform fuzzy processing on the preprocessed data to extract objects of interest, including: A fuzzy clustering algorithm based on dynamic thresholds is used to perform fuzzy processing on the preprocessed data to obtain a fuzzy membership matrix that represents the membership relationship between objects and laser beams. The number of objects within the field of view of the lidar is determined by the fuzzy membership matrix, as well as the lidar segment group corresponding to the object. Then, the distance to the object is obtained according to the lidar ranging formula. The cluster closest to the lidar sensor is selected as the object of interest; Also includes: An improved particle swarm optimization algorithm is used to dynamically adjust the initial parameters of the adaptive fuzzy PID controller with minimum error. The particle update strategy of the improved particle swarm optimization algorithm is as follows: in, Indicates the first The position of each particle in the current search space. This represents the proportional gain parameter. This represents the integral gain parameter. Represents the differential gain parameter. This indicates the time step range for error calculation. Indicates in Error at the next iteration This represents the objective function that needs to be minimized. Indicates the first The particle in the first Speed ​​at the next iteration Represents the inertia weight term. Indicates the first The current optimal position found by each particle. and Represents a random number between [0,1]. Indicates the first learning factor. Indicates the second learning factor. This represents the globally optimal position found among all particles. particle exist Position at the next iteration and These represent the maximum and minimum values ​​of the inertia weight, respectively. Indicates the number of iterations. Maximum number of iterations They represent and The maximum value, They represent and The minimum value.

2. The method for coordinated control of heavy-haul train groups according to claim 1, characterized in that, Real-time segmented data, including: The relative distance between the object of interest and the heavy-duty train equipped with lidar, the speed of the object of interest relative to the heavy-duty train equipped with lidar, and the segment distance of the solid-state lidar. Among them, the relative distance between the object of interest and the heavy-haul train equipped with lidar and the speed of the object of interest relative to the heavy-haul train equipped with lidar are directly measured; The segment distance D of the solid-state lidar is calculated using the following formula: in, This indicates the segmented distance of the lidar. and These represent the distances between the two laser beams emitted by the solid-state lidar. This indicates the angle between the two laser beams emitted by a solid-state lidar. Indicates angle cosine value, This indicates the number of segments in the lidar system.

3. The method for coordinated control of heavy-haul train groups according to claim 1, characterized in that, An adaptive fuzzy clustering algorithm based on dynamic thresholds is defined by the following formula: Where σ represents the σ-th iteration, Represents the first group of the pregrouped matrix Column, i.e., the first Segmentation, Represents the first group of the pregrouped matrix Line, i.e., the first Segmentation group, Denotes the σ-th iteration. The center position of the segment group Represents the pregrouping after the σ-th iteration. The first segment group The degree of belonging to the segment, D l Indicates the current number l Segmented lidar for segmented distance measurement. Represents the first group of the pregrouped matrix The first segment group Segmented lidar range This indicates the number of segments in the lidar system. This represents the optimal number of segments. Describes the σ-th iteration of the binary matrix. The first segment group The degree of belonging to each segment Represents the fuzzy index. Iteration The next Changes in the center of the segment group Indicates the initial threshold. The convergence threshold represents the dynamic change of the σ-th iteration. This represents the error or change value in the σ-th iteration. Used to measure the state of the algorithm in the σ-th iteration. This indicates a regulatory factor.

4. The method for coordinated control of heavy-haul train groups according to claim 1, characterized in that, Adjusting the distance and speed between the following train and the heavy-haul train ahead, including: The distance error value is calculated based on the distance between the following train and the preceding heavy-haul train, and the expected cooperative distance. The speed of the heavy-haul train is then adjusted using cooperative control based on this distance error value. The cooperative control adjustment is expressed as follows: in, This represents the three-dimensional distance to the object closest to the lidar. Indicates the expected collaborative distance. This indicates the distance traveled by the following train when there is a communication delay with adjacent trains. Indicates the maximum braking distance of the following vehicle. Indicates the maximum braking distance of the vehicle in front. This indicates a safety margin, and max() indicates taking the maximum value. Indicates time, This represents the error between the actual distance of the train and the target distance at time t. Indicates the train's time. The actual collaborative distance Indicates error rate of change, Indicates the sampling period. This represents the error at the previous sampling time. Indicates the train's time. speed, Indicates the expected speed of collaboration. Obtained from driving map This represents the gain coefficient, which controls the degree of speed response to distance errors. Indicates the distance of idling and stopping. This indicates the common braking distance of the train operation monitoring and recording device LKJ. Indicates the initial braking speed. Indicates the braking calculation coefficient. Indicates the train's switching braking rate. This represents the friction coefficient converted from brake shoes. Indicates common braking parameters, This represents the basic unit resistance of a train. Indicates the driver's reaction time. This indicates the final speed of a heavily loaded train after it has slowed down.

5. The method for coordinated control of heavy-haul train groups according to claim 4, characterized in that, The adaptive fuzzy PID controller that minimizes error is formulated as follows: in, This represents the output signal of the controller at time t. This represents the error between the actual distance of the train and the target distance at time t. Indicates time, Indicates error rate of change, This represents the proportional gain parameter. This represents the integral gain parameter. Represents the differential gain parameter. Indicates at time proportional gain, Indicates time The differential gain, Indicates at time Integral gain, This represents the initial proportional gain. Indicates the initial integral gain. This represents the initial differential gain. This indicates the proportional gain adjustment amount. This represents the differential gain adjustment amount. This indicates the integral gain adjustment amount. , , The three adjustment values ​​were obtained through fuzzy reasoning.

6. A coordinated control device for heavy-haul train groups, characterized in that, include: The system comprises a segmented data acquisition unit, a preprocessing unit, an object of interest extraction unit, and a collaborative control unit. The segmented data acquisition unit is used to acquire real-time segmented data of the surrounding environment of the heavy-haul train group based on lidar in the following train in the heavy-haul train group. The preprocessing unit is used to preprocess real-time segmented data using dynamic programming to obtain preprocessed data. The object of interest extraction unit is used to perform fuzzy processing on the preprocessed data using a fuzzy clustering algorithm based on dynamic thresholds, extract the object of interest, and obtain the distance between the following vehicle and all objects of interest within a preset distance range; The cooperative control unit is used to adjust the distance and speed between the following vehicle and the heavy-load train ahead based on the distance between the following vehicle and the heavy-load train ahead, using an adaptive fuzzy PID controller that minimizes errors, when there is only a heavy-load train ahead within a preset distance range. The preprocessing unit is specifically used for: The collected lidar segmented data is preprocessed using dynamic programming to generate the corresponding binary membership matrix as preprocessed data. The formula for dynamic programming is as follows: in, This represents the distance data between segments of a solid-state lidar system. This indicates the number of segments in the lidar system. The first laser radar Segmented distance data, Indicates the first The difference between the m-th and m-th adjacent data points, where diff[m] represents the difference between the m-th and m-th adjacent data points. This represents the starting index of the lidar range data segment, where dp[i] indicates the previous segment. The optimal grouping cost for segmentation, dp[j] represents the cost of the first segment. The optimal grouping cost for segmentation, cost(j+1,i) represents the cost from segmentation. arrive The cost and, This represents the optimal split point that minimizes the total cost. Let j represent the set of all optimal cut points, and j1 represent the optimal cut point of the first segment group. Indicates the first Optimal cutting point of segment group This represents the optimal number of segments. Indicates the first The optimal cutting point express The starting position of the segment; | | represents the absolute value of P; This represents the value of the variable that minimizes the expression, i.e., the value of the optimal segmentation point; the formula for the binary membership matrix is ​​as follows: in, Indicates the first The first in the segment Each data point is an element in the data matrix. This indicates that the current number is the [number]. Segmentation, Indicates the first data point in the current data point sequence. One element, This represents an element in the data sequence, specifically the first element in the input data. Data points, Indicates the first The first in the segment Each binary affiliation matrix element; The object of interest extraction unit is specifically used for: A fuzzy clustering algorithm based on dynamic thresholds is used to perform fuzzy processing on the preprocessed data to obtain a fuzzy membership matrix that represents the membership relationship between objects and laser beams. The number of objects within the field of view of the lidar is determined by the fuzzy membership matrix, as well as the lidar segment group corresponding to the object. Then, the distance to the object is obtained according to the lidar ranging formula. The cluster closest to the lidar sensor is selected as the object of interest; An improved particle swarm optimization algorithm is used to dynamically adjust the initial parameters of the adaptive fuzzy PID controller with minimum error. The particle update strategy of the improved particle swarm optimization algorithm is as follows: in, Indicates the first The position of each particle in the current search space. This represents the proportional gain parameter. This represents the integral gain parameter. Represents the differential gain parameter. This indicates the time step range for error calculation. Indicates in Error at the next iteration This represents the objective function that needs to be minimized. Indicates the first The particle in the first Speed ​​at the next iteration Represents the inertia weight term. Indicates the first The current optimal position found by each particle. and Represents a random number between [0,1]. Indicates the first learning factor. Indicates the second learning factor. This represents the globally optimal position found among all particles. particle exist Position at the next iteration and These represent the maximum and minimum values ​​of the inertia weight, respectively. Indicates the number of iterations. Maximum number of iterations They represent and The maximum value, They represent and The minimum value.

7. A heavy-haul train group coordinated control device according to claim 6, characterized in that, Real-time segmented data, including: The distance from the object of interest to the lidar, the speed of the object of interest relative to the heavy-duty train carrying the lidar, and the segment distance of the solid-state lidar; Among them, the distance from the object of interest to the lidar and the speed of the object of interest relative to the heavy-duty train equipped with the lidar are directly measured by the solid-state lidar; The segment distance D of the solid-state lidar is calculated using the following formula: in, This indicates the segmented distance of the lidar. and These represent the distances between the two laser beams emitted by the solid-state lidar. This indicates the angle between the two laser beams emitted by a solid-state lidar. Indicates angle cosine value, This indicates the number of segments in the lidar system.

8. A heavy-haul train group coordinated control device according to claim 6, characterized in that, An adaptive fuzzy clustering algorithm based on dynamic thresholds is defined by the following formula: Where σ represents the σ-th iteration, Represents the first group of the pregrouped matrix Column, i.e., the first Segmentation, Represents the first group of the pregrouped matrix Line, i.e., the first Segmentation group, Denotes the σ-th iteration. The center position of the segment group Represents the pregrouping after the σ-th iteration. The first segment group The degree of belonging to the segment, D l Indicates the current number l Segmented lidar for segmented distance measurement. Represents the first group of the pregrouped matrix The first segment group Segmented lidar range This indicates the number of segments in the lidar system. This represents the optimal number of segments. Describes the σ-th iteration of the binary matrix. The first segment group The degree of belonging to each segment Represents the fuzzy index. Iteration The next Changes in the center of the segment group Indicates the initial threshold. The convergence threshold represents the dynamic change of the σ-th iteration. This represents the error or change value in the σ-th iteration. Used to measure the state of the algorithm in the σ-th iteration. This indicates a regulatory factor.

9. A heavy-haul train group coordinated control device according to claim 6, characterized in that, The collaborative control unit is specifically used for: The distance error value is calculated based on the distance between the following train and the preceding heavy-haul train, and the expected cooperative distance. The speed of the heavy-haul train is then adjusted using cooperative control based on this distance error value. The cooperative control adjustment is expressed as follows: in, This represents the three-dimensional distance to the object closest to the lidar. Indicates the expected collaborative distance. This indicates the distance traveled by the following train when there is a communication delay with adjacent trains. Indicates the maximum braking distance of the following vehicle. Indicates the maximum braking distance of the vehicle in front. This indicates a safety margin, and max() indicates taking the maximum value. Indicates time, This represents the error between the actual distance of the train and the target distance at time t. Indicates the train's time. The actual collaborative distance Indicates error rate of change, Indicates the sampling period. This represents the error at the previous sampling time. Indicates the train's time. speed, Indicates the expected speed of collaboration. Obtained from driving map This represents the gain coefficient, which controls the degree of speed response to distance errors. Indicates the distance of idling and stopping. This indicates the common braking distance of the train operation monitoring and recording device LKJ. Indicates the initial braking speed. Indicates the braking calculation coefficient. Indicates the train's switching braking rate. This represents the friction coefficient converted from brake shoes. Indicates common braking parameters, This represents the basic unit resistance of a train. Indicates the driver's reaction time. This indicates the final speed of a heavily loaded train after it has slowed down.

10. A heavy-haul train group coordinated control device according to claim 9, characterized in that, The adaptive fuzzy PID controller that minimizes error is formulated as follows: in, This represents the output signal of the controller at time t. This represents the error between the actual distance of the train and the target distance at time t. Indicates time, Indicates error rate of change, This represents the proportional gain parameter. This represents the integral gain parameter. Represents the differential gain parameter. Indicates at time proportional gain, Indicates time The differential gain, Indicates at time Integral gain, This represents the initial proportional gain. Indicates the initial integral gain. This represents the initial differential gain. This indicates the proportional gain adjustment amount. This represents the differential gain adjustment amount. This indicates the integral gain adjustment amount. , , The three adjustment values ​​were obtained through fuzzy reasoning.

11. An electronic device, characterized in that, It includes a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus; Memory, which stores computer programs; A processor, when executing a computer program stored in memory, implements a heavy-haul train group cooperative control method according to any one of claims 1-5.

12. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements a heavy-load train group cooperative control method according to any one of claims 1-5.

Images