Heavy-load train group operation control method based on enhanced Kalman filtering estimation

Abstract

The invention relates to the technical field of rail transit heavy haul train control, in particular to a heavy haul train group operation control method based on enhanced Kalman filtering estimation. The method comprises the following steps: acquiring an error covariance matrix at the current moment; determining estimated state information of the adjacent train based on the error covariance matrix at the current moment; after the speed of the current train is controlled based on the state information of the adjacent train, whether delay state information of the adjacent train is received or not is judged; if yes, controlling the speed of the current train based on the delay state information of the adjacent train; and if not, updating the error covariance matrix at the current moment based on the pre-estimated state information of the adjacent train, and then re-executing the step of determining the pre-estimated state information of the adjacent train based on the error covariance matrix at the current moment. Therefore, the running stability and safety of the heavy haul train group can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of rail transit heavy-load train control, and in particular to a heavy-load train group operation control method based on enhanced Kalman filter estimation. Background Art

[0002] Railways are the main arteries of the national economy, key infrastructure, and important livelihood projects. They are the backbone of the comprehensive transportation system and an important mode of transportation, and are of vital importance to my country's economic and social development. In recent years, with the growth of energy demand, heavy-load railway coal transportation has faced a severe capacity shortage. To improve transportation efficiency, the industry generally adopts the method of lengthening train formations and increasing axle loads. However, such measures complicate the longitudinal dynamic characteristics of trains, easily leading to problems such as uneven force on couplers and insufficient coordination of braking systems. This results in frequent acceleration and deceleration during operation, which not only reduces energy efficiency but also increases equipment wear and safety risks.

[0003] Against this backdrop, the heavy-haul train group operation model has emerged, enhancing overall transport capacity through coordinated multi-train formation. However, due to the strong dynamic coupling between trains within a group and the numerous interfering factors in the operating environment (such as grade changes and wind resistance fluctuations), traditional control methods, which rely on fixed models or single Kalman filter estimation, have difficulty accurately tracking train status. This leads to delayed control commands and frequent speed adjustments, compromising the stability and safety of group operations.

[0004] Based on this, the present invention proposes a heavy-load train group operation control method based on enhanced Kalman filter estimation to solve the above technical problems. Summary of the Invention

[0005] The present invention describes a heavy-load train group operation control method based on enhanced Kalman filter estimation, which can improve the stability and safety of heavy-load train group operation.

[0006] According to a first aspect, the present invention provides a method for controlling the operation of a heavy-load train group based on enhanced Kalman filter estimation, the method comprising:

[0007] When the communication between the current train and the adjacent train in the heavy-load train group is interrupted, the error covariance matrix at the current moment is obtained;

[0008] Determining estimated state information of adjacent trains based on the error covariance matrix at the current moment;

[0009] After controlling the speed of the current train based on the status information of the adjacent train, determining whether the delay status information of the adjacent train has been received;

[0010] If so, controlling the speed of the current train based on the delay status information of the adjacent train;

[0011] If not, after updating the error covariance matrix at the current moment based on the estimated state information of the adjacent train, re-execute the step of "determining the estimated state information of the adjacent train based on the error covariance matrix at the current moment".

[0012] According to a second aspect, the present invention provides a heavy-load train group operation control device based on enhanced Kalman filter estimation, comprising:

[0013] an acquisition unit configured to acquire an error covariance matrix at a current moment when communication between a current train and an adjacent train in the heavy-load train group is interrupted;

[0014] a first data processing unit configured to determine estimated state information of adjacent trains based on the error covariance matrix at the current moment;

[0015] a second data processing unit configured to, after controlling the speed of the current train based on the status information of the adjacent train, determine whether the delay status information of the adjacent train has been received;

[0016] a third data processing unit, configured to, if yes, control the speed of the current train based on the delay status information of the adjacent train;

[0017] The fourth data processing unit is configured to, if not, update the error covariance matrix at the current moment based on the estimated state information of the adjacent train, and then re-execute the step of "determining the estimated state information of the adjacent train based on the error covariance matrix at the current moment".

[0018] According to the enhanced Kalman filter estimation-based heavy-haul train group operation control method provided by the present invention, in heavy-haul train group operation scenarios, the stability of the inter-train communication link directly affects operational safety and efficiency. When communication is interrupted between the current train and an adjacent train in the group, an emergency control mechanism is immediately activated: first, the current error covariance matrix is ​​acquired in real time. This matrix quantifies the uncertainty of the adjacent train's state estimate. Based on this matrix, the estimated state information of the adjacent train is determined, providing data support for control decisions. The current train's speed is simultaneously regulated based on the adjacent train's state information, while the communication link is continuously monitored to determine whether delayed state information from the adjacent train has been received. If delayed data is received, the latest state information is integrated into the current train's speed for dynamic compensation control, ensuring that the control strategy is synchronized with the actual operating state. If not, the error covariance matrix is ​​updated based on the estimated state information of the adjacent train. After optimizing the state prediction accuracy, the state prediction process is re-executed, forming a closed-loop feedback control. This adaptive emergency mechanism effectively addresses the control lag caused by communication interruptions and significantly improves the stability and safety of heavy-haul train group operations. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0020] Figure 1 A flow chart of a method for controlling the operation of a heavy-load train group based on enhanced Kalman filter estimation according to one embodiment is shown;

[0021] Figure 2 A schematic block diagram of a heavy-load train group operation control device based on enhanced Kalman filter estimation according to one embodiment is shown;

[0022] Figure 3 A schematic diagram of heavy-load train group operation control based on enhanced Kalman filter estimation according to one embodiment is shown;

[0023] Figure 4 A flowchart of heavy-load train group operation control based on enhanced Kalman filter estimation according to one embodiment is shown. DETAILED DESCRIPTION

[0024] The solution provided by the present invention is described below with reference to the accompanying drawings.

[0025] Figure 1 The flowchart of the heavy-load train group operation control method based on enhanced Kalman filter estimation according to one embodiment is shown. It can be understood that the method can be executed by any device, equipment, platform, or equipment cluster with computing and processing capabilities. Figure 1 As shown, the method includes:

[0026] Step 100: When the communication between the current train and the adjacent train in the heavy-load train group is interrupted, obtain the error covariance matrix at the current moment;

[0027] Step 102: Determine the estimated state information of the adjacent train based on the error covariance matrix at the current moment;

[0028] Step 104: After controlling the speed of the current train based on the status information of the adjacent train, determine whether the delay status information of the adjacent train has been received;

[0029] Step 106: If yes, control the speed of the current train based on the delay status information of the adjacent train;

[0030] Step 108: If not, after updating the error covariance matrix at the current moment based on the estimated state information of the adjacent train, re-execute the step of "determining the estimated state information of the adjacent train based on the error covariance matrix at the current moment".

[0031] In this embodiment, in heavy-haul train group operation scenarios, the stability of inter-train communication links directly impacts operational safety and efficiency. When communication is interrupted between the current train and an adjacent train in the group, an emergency control mechanism is immediately activated. First, the current error covariance matrix is ​​acquired in real time. This matrix quantifies the uncertainty of the adjacent train's state estimate. Based on this matrix, the estimated state information of the adjacent train is determined, providing data support for control decisions. The current train's speed is then regulated based on the adjacent train's state information. Meanwhile, the communication link is continuously monitored to determine whether delayed state information from the adjacent train has been received. If delayed data is received, the latest state information is integrated into the current train's speed for dynamic compensation control, ensuring that the control strategy is synchronized with the actual operating state. If not, the error covariance matrix is ​​updated based on the estimated state information from the adjacent train. The state prediction process is then re-executed after optimizing the state prediction accuracy, forming a closed-loop feedback control. This adaptive emergency mechanism effectively addresses the control lag caused by communication interruptions and significantly improves the stability and safety of heavy-haul train group operations.

[0032] In one embodiment of the present invention, the estimated state information of adjacent trains is determined by the following formula:

[0033] x i (t k )=A e,i x i (t k-1 )+B e,i u i (t k-1 )+w e,i (t k-1 )

[0034] y i (t k )=C e,i x i (t k-1 )+ν e,i (t k-1 )

[0035]

[0036] Where i is the number of the adjacent train, t k is the current moment, x i (t k )=[s i (tk ),v i (t k ),a i (t k )] T For t k The status information of train i in the heavy-load train group, s i (t k ) is the train i in the heavy-load train group at t k The position at the moment, v i (t k ) is the train i in the heavy-load train group at t k The speed of the moment, a i (t k ) is the train i in the heavy-load train group at t k Acceleration at the moment, u i (t k-1 ) is t k-1 The control input of the actuator of train i at time A e,i and B e,i is the known state matrix and control matrix of train i, w e,i (t k ) is the train i at t k The process noise at the moment, R e is the process noise covariance matrix, y i (t k ) is t k The observed state information of train i at time C e,i =[1,0,0] T is the observation matrix, ν e,i (t k ) is the time during which train i runs k The measurement noise at the moment, Ne is the covariance matrix of the measurement noise, t k The estimated state information of train i before correction obtained by the enhanced Kalman filter estimation algorithm at time t, that is, the estimated state information of the adjacent train, At time t k-1 The estimated state information of train i in the heavy-load train group is The train i at t is obtained by the enhanced Kalman filter estimation method k The error covariance matrix at time , t k The Kalman gain of train i at time instant, C e,i is the observation matrix, C e,i T C e,i The transposed matrix of At time t kThe estimated state information of train i in the heavy-load train group, χ i (t k ) is the train i at t k The error covariance matrix updated at the moment, that is, the error covariance matrix at the current moment, χ i (t k-1 ) is the train i at t k-1 The error covariance matrix updated at all times, I is the identity matrix.

[0037] In this embodiment, when the virtual fleet encounters interference, the state information of train i can be expressed as:

[0038]

[0039] Among them, x i (t k ) indicates that at time t k The status information of train i in the heavy-load train group is Indicates that at time t k The estimated value of the state information of train i in the heavy-load train group.

[0040] like Figure 3 As shown, in this embodiment, after the heavy-load train group is unmarshaled, it relies on ZC to send the MA information of the train operation in the current state. Similarly, the present invention defines the Kalman filter estimation process as a discrete control process, and the random state control matrix of train i is shown in formula (2):

[0041] x i (t k )=A e,i x i (t k-1 )+B e,i u i (t k-1 )+w e,i (t k-1 ) (2)

[0042] Where x i (t k )=[s i (t k ),v i (t k ),a i (t k )] T Indicates that at time t k The status information of train i in the heavy-load train group is s i (t k ) represents the train i in the heavy-load train group at t k Position at the moment; v i(t k ) represents the train i in the heavy-load train group at t k The speed of time; a i (t k ) represents the train i in the heavy-load train group at t k Acceleration at the moment; u i (t k-1 ) is t k-1 The control input of the actuator of train i at time A e,i and B e,i is the known state matrix and control matrix of train i, and the specific parameters vary depending on the vehicle type; w e,i (t k ) represents train i at t k The process noise at the moment obeys the independent distribution of the Gaussian matrix based on the zero-sum distribution, that is, w e,i (t k )~(0,R e ), and they all obey independent zero-sum distributed Gaussian matrices, R e is the process noise covariance matrix.

[0043] At the same time, the observation equation of the enhanced Kalman filter is as follows:

[0044] y i (t k )=C e,i x i (t k-1 )+v e,i (t k-1 ) (3)

[0045] Where y i (t k ) is t k The observed status information of train i at time C e,i =[1,0,0] T is the observation matrix; ν e,i (t k ) is the time during which train i runs k The measurement noise at the moment obeys the independent distribution of the Gaussian matrix based on the zero-sum distribution, that is, ν e,i (t k )~(0,N e ), Ne is the covariance matrix of measurement noise.

[0046] To better evaluate the system performance of a heavy-load train group under interference, the present invention divides its performance indicators into two layers: the prediction side based on the transmission update time and the reception information side based on the measurement update. The prediction formula of the enhanced Kalman filter estimation algorithm is as follows:

[0047]

[0048] in, t k The state information prediction value of train i before correction obtained by the enhanced Kalman filter estimation algorithm at time t (error correction is required later using formula 6); Indicates that at time t k-1 The estimated value of the state information of train i in the heavy-load train group when A is e,i and B e,i is the known state matrix and control matrix of train i, and the specific parameters vary depending on the vehicle type; u i (t k-1 ) is t k-1 The control input of the actuator of train i at time t is estimated by the enhanced Kalman filter algorithm. k-1 The estimated state information of the preceding vehicle obtained at any moment The calculation is the original calculation part of the system, and the present invention does not provide a separate calculation method;

[0049] Furthermore, by combining the optimal interference strategy with the enhanced Kalman filter algorithm, an enhanced Kalman filter train state information estimation method is proposed to alleviate the situation where the train cannot receive the state information of the preceding train during operation. The prediction of the state information of the heavy-load train relies on the update of the measurement information during operation. Therefore, the data update process during the train operation can be defined as shown in Equation (5-7):

[0050]

[0051] Where, The train i at t is obtained by the enhanced Kalman filter estimation method k The error covariance matrix at time , t k The Kalman gain of train i at time i; C e,i is the observation matrix; C e,i T C e,i The transposed matrix of R e is the process noise covariance matrix; Indicates that at time t k The estimated value of the state information of train i in the heavy-load train group; t k The state information prediction value of train i before correction obtained by the enhanced Kalman filter estimation algorithm at time y; i (t k ) is t k The observed state information of train i at time χ i(t k ) is the heavy-load train i at t k The error covariance matrix updated at all times; χ i (t k-1 ) is the heavy-load train i at t k-1 The error covariance matrix updated at all times; I is the identity matrix.

[0052] In one embodiment of the present invention, the k The status information of train i in the heavy-load train group is expressed by the following formula:

[0053] x(t k ):=[s(t v ), v(t k ), a(t k )] T

[0054] s(t k ):=[s1(t k ), s2(t k ),…,s n (t k )] T

[0055] v(t k ):=[v1(t k ), v2(t k ),…,ν n (t k )] T

[0056] a(t k ):=[a1(t k ), a2(t k ),…,a n (t k )] T

[0057] In the formula, x(t k )=[x1(t k ),x2(t k ),...,x n (t k )] T ,x(t k ) indicates that at time t k Status information of the heavy-load train group, s i (t k ) represents the train i in the heavy-load train group at t k The position at the moment, v i (t k ) represents the train i in the heavy-load train group at tk The speed of the moment, a i (t k ) represents the train i in the heavy-load train group at t k The acceleration at the moment, n is the number of trains in the heavy-load train group.

[0058] In this embodiment, the algorithm of the present invention is implemented on the traction locomotive of a heavy-load train, and its state information is defined as: position, speed and acceleration information during the heavy-load train line operation, and its specific expression is shown in the formula.

[0059] x(t k ):=[s(t k ), v(t k ), a(t k )] T (8)

[0060] s(t k ):=[s1(t k ), s2(t k ),...,s n (t k )] T (9)

[0061] v(t k ):=[v1(t k ), v2(t k ), ..., v n (t k )] T (10)

[0062] a(t k ):=[a1(t k ), a2(t k ), ..., a n (t k )] T (11)

[0063] Among them, x(t k )=[x1(t k ),x2(t k ),...,x n (t k )] T ,x(t k ) indicates that at time t k Status information of the heavy-load train group, s i (t k ) represents the train i in the heavy-load train group at t k The position at the moment, v i (t k) represents the train i in the heavy-load train group at t k The speed of the moment, a i (t k ) represents the train i in the heavy-load train group at t k The acceleration at the moment, n is the number of trains in the heavy-load train group.

[0064] Next, the control model of the heavy-load train group is transformed into a linear system. The mathematical expression of the entire train is as follows:

[0065] x i+1 (t k )=A e,(i+1) x i+1 (t k-1 )+B e,i+1 u i+1 (t k )+w e,i+1 (t k ) (12)

[0066] Among them, x i+1 (t k )=[s i+1 (t k ),v i+1 (t k ),a i+1 (t k )] T is the status information of train i+1; u i+1 (t k ) is the control input of the train i+1 controller, which is estimated by the enhanced Kalman filter algorithm at t k The estimated state information of the preceding vehicle obtained at any moment It is calculated that this part is the original calculation part of the system, and the present invention will no longer provide a separate calculation method.

[0067] In one embodiment of the present invention, the error covariance matrix at the current moment is updated using the following formula:

[0068]

[0069] Where, is the minimum data packet success rate allowed by the system, χ i - (t k ) is t k The estimated error covariance matrix of the state information of train i at time t is the updated error covariance matrix, χ i - (t k-1 ) is t k-1The estimated error covariance matrix of the state information of train i at time t is the error covariance matrix at the current time, A e,i is the known state matrix of train i, M e is a matrix of known dimensions, R e is the process noise covariance matrix.

[0070] In this embodiment, generally speaking, interference reduces the SINR value of the current communication by sending the same spectrum as the channel to the communication link between the interference target train and the base station, resulting in packet loss, delay, and a decrease in the successful reception rate of data packets. Furthermore, by affecting the vehicle-to-vehicle / vehicle-to-ground communication link, the trains in the heavy-load train group cannot receive the status information of the adjacent trains in real time within the allowed time period, thereby paralyzing the vehicle-to-vehicle communication system. Therefore, communication interference will ultimately lead to random communication delays and random de-grouping of the heavy-load train group. This chapter proposes a new train state estimation algorithm to improve the system stability during the interference process. The present invention defines communication interference and algorithm estimation error as two parties in the game and designs a cost function as one of the indicators to measure the benefits of both parties. By optimizing the train state information estimation error to minimize it, the accuracy of the control system is achieved, so that the heavy-load train group can still operate stably even when the vehicle-to-vehicle communication and vehicle-to-ground communication are unreliable. In order to minimize the estimation error, the present invention first designs the state information estimation error covariance matrix of train i, as shown in formula (13).

[0071]

[0072] Among them, χ i - (t k ) is t k The covariance matrix of the estimated error of the state information of train i at time instant; E[·] is the expectation function; t k The estimated error of the state information of train i at time; for The transposed matrix of .

[0073] Furthermore, the state information estimation error The mathematical expression of is shown in formula (14).

[0074]

[0075] Where, t k The estimated error of the state information of train i at time x i (t k )=[s i (t k ),v i (tk ),a i (t k )] T is the status information of train i; t k Estimated value of the status information of train i at time.

[0076] Furthermore, in order to minimize the covariance of the state information estimation error, the objective function under communication interference conditions is designed as shown in formula (16):

[0077]

[0078] Where, The train i at t is obtained by the enhanced Kalman filter estimation method k The error covariance matrix at time , Indicates that at t k The energy consumed by the communication interference sent to the vehicle-to-vehicle communication channel at all times; Indicates that at t k Assume that the maximum energy consumption required to interfere with the vehicle-to-vehicle communication channel due to communication interference at time T1 and T2 is the start and end time of the interference, respectively; ΔT = T2 - T1 is the duration of the communication interference. This paper explores the factors affecting communication on heavy-load train formations. SINR (Signal-to-Interference-and-Noise Ratio) and packet transmission success rate are key indicators for evaluating wireless communication system performance. This paper primarily uses SINR as a measure of channel quality.

[0079] To facilitate subsequent derivation, the channel data packet transmission success rate is defined as follows:

[0080]

[0081] Where, f e (t k ) is the success rate of data packet transmission through the channel, which is about and A continuous function of t k The signal attenuation of the channel at time t; Indicates that at t k The energy consumed by the communication interference sent into the vehicle-to-vehicle communication channel is assumed at all times.

[0082] In order to better discuss the relationship between communication interference and packet loss rate, according to the principles of digital communication, the conversion relationship between bit error rate (Symbol Error Rate, SER) and SINR can be expressed as the following formula:

[0083]

[0084] Where, is the channel bit error rate; t k SINR value at the moment; is a constant parameter; t k The transmit power of the eNodeB / antenna at the time; ω e is the channel noise; In e Communication interference exists in the channel; t k The signal attenuation of the channel at time t; Indicates that at t k The energy consumed by the communication interference sent into the vehicle-to-vehicle communication channel is assumed at all times.

[0085] In order to facilitate the evaluation of the system performance of the network layer of the heavy-load train group system, the data packet transmission success rate of the system under communication interference is shown in the following formula:

[0086]

[0087] Where, f e (t k ) is the success rate of data packet transmission through the channel, which is about and A continuous function of t k The signal attenuation of the channel at time t; Indicates that at t k The energy consumed by the communication interference sent to the vehicle-to-vehicle communication channel at all times; is a constant parameter; t k SINR value at the moment.

[0088] By deducing and solving, we can obtain the following solution:

[0089]

[0090] Where, is the energy function consumed by the calculated channel interference; e is a known constant; t k The transmit power of the eNodeB / antenna at that moment; t k The signal attenuation of the channel at time t; t k SINR value at the moment; ω e is the channel noise; In e The communication interference existing in the channel.

[0091] The energy strategy for communication interference can be estimated by derivation as follows:

[0092]

[0093] Where, t k The transmit power of the eNodeB / antenna at that moment; t k The signal attenuation of the channel at time t; t k SINR value at the moment; ω e is the channel noise; In e Communication interference exists in the channel; t k The packet loss rate of the vehicle-to-vehicle communication channel at time M e is a matrix of known dimensions, which is given by the following formula. Based on the above analysis, it can be seen that the estimated error covariance matrix is ​​adjusted in real time by iteratively updating in each communication cycle. The matrix M e and train i in t k The error covariance matrix at time The result can be obtained by solving the Lyapunov function, as shown in the following formula:

[0094]

[0095] Where, is the minimum data packet success rate allowed by the system; i - (t k ) is t k The covariance matrix of the estimated error of the state information of train i at time x i - (t k-1 ) is t k-1 The covariance matrix of the estimated error of the state information of train i at time A e,i is the known state matrix of train i, and its specific parameters vary depending on the type of train; M e For a matrix of known dimensions, it needs to be calculated using formula (18); R e is the process noise covariance matrix.

[0096] like Figure 4As shown, in this embodiment, the process begins with the state information exchange link of the marshaling train. The system first detects whether it has received the real-time position and speed information of the adjacent train (step S1). When the communication link is normal, multi-vehicle dynamic coupling modeling is performed using the observation equation (step S2), directly obtaining the precise motion parameters of the adjacent train. If communication is interrupted or interfered with, the enhanced Kalman filter estimation mechanism is activated (step S3) to generate compensation information for the adjacent train state. This process specifically designs an error tolerance mechanism to address the uncertainty of wireless communication in virtual marshaling. For adjacent train data with transmission delays (step S4), the system dynamically corrects it using the energy function optimization module (step S5). Simultaneously, an error matrix minimization operation is performed in the case of undelayed data (step S6). These two operations jointly ensure the accuracy of state fusion under different communication conditions. After completing the information processing, the algorithm calculates the acceleration command of the adjacent train and the dynamic distance between the adjacent train and the adjacent train based on the train dynamics model (step S7). The acceleration control variable is related to key parameters such as line grade and train mass through a nonlinear function. The safety verification process employs dual constraints: first, verifying whether the current speed exceeds the line's dynamic limit curve (step S8). If an overspeed risk is detected, graded braking is immediately triggered. Second, the distance between trains in the train formation is monitored in real time (step S9). If the minimum safety threshold is exceeded, emergency braking is initiated for the entire train (step S10). The resulting acceleration command is then implemented by the electronically controlled actuator (step S11), and the updated state parameters of the train are transmitted to other trains in the formation via a wireless transmission system, forming a closed-loop control circuit.

[0097] The foregoing description describes specific embodiments of the present invention. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims can be performed in an order different from that described in the embodiments and still achieve the desired results. Furthermore, the processes depicted in the accompanying drawings do not necessarily require the specific order shown or the sequential order to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0098] According to another embodiment, the present invention provides a heavy-load train group operation control device based on enhanced Kalman filter estimation. Figure 2 A schematic block diagram of a heavy-load train group operation control device based on enhanced Kalman filter estimation according to one embodiment is shown. It can be understood that the device can be implemented by any device, equipment, platform and equipment cluster with computing and processing capabilities. Figure 2 As shown, the device includes: an acquisition unit 200, a first data processing unit 202, a second data processing unit 204, a third data processing unit 206, and a fourth data processing unit 208. The main functions of each component unit are as follows:

[0099] The acquisition unit 200 is configured to acquire the error covariance matrix at the current moment when communication between the current train and the adjacent train in the heavy-load train group is interrupted;

[0100] A first data processing unit 202 is configured to determine estimated state information of adjacent trains based on the error covariance matrix at the current moment;

[0101] The second data processing unit 204 is configured to control the speed of the current train based on the status information of the adjacent train and then determine whether the delay status information of the adjacent train is received;

[0102] The third data processing unit 206 is configured to, if yes, control the speed of the current train based on the delay status information of the adjacent train;

[0103] The fourth data processing unit 208 is configured to, if not, update the error covariance matrix at the current moment based on the estimated state information of the adjacent train, and then re-execute the step of "determining the estimated state information of the adjacent train based on the error covariance matrix at the current moment".

[0104] In one embodiment of the present invention, the estimated state information of the adjacent train is determined by the following formula:

[0105] x i (t k )=A e,i x i (t k-1 )+B e,i u i (t k-1 )+w e,i (t k-1 )

[0106] y i (t k )=C e,i x i (t k-1 )+v e,i (t k-1 )

[0107]

[0108] Where i is the number of the adjacent train, t k is the current moment, x i (t k )=[s i (t k ),v i (t k ),a i (tk )] T For t k The status information of train i in the heavy-load train group, s i (t k ) is the train i in the heavy-load train group at t k The position at the moment, v i (t k ) is the train i in the heavy-load train group at t k The speed of the moment, a i (t k ) is the train i in the heavy-load train group at t k Acceleration at the moment, u i (t k-1 ) is t k-1 The control input of the actuator of train i at time A e,i and B e,i is the known state matrix and control matrix of train i, w e,i (t k ) is the train i at t k The process noise at the moment, R e is the process noise covariance matrix, y i (t k ) is t k The observed state information of train i at time C e,i =[1,0,0] T is the observation matrix, ν e,i (t k ) is the time during which train i runs k The measurement noise at the moment, Ne is the covariance matrix of the measurement noise, t k The estimated state information of train i before correction obtained by the enhanced Kalman filter estimation algorithm at time t, that is, the estimated state information of the adjacent train, At time t k-1 The estimated state information of train i in the heavy-load train group is The train i at t is obtained by the enhanced Kalman filter estimation method k The error covariance matrix at time , t k The Kalman gain of train i at time instant, C e,i is the observation matrix, C e,i T C e,i The transposed matrix of At time t k The estimated state information of train i in the heavy-load train group, χ i (t k ) is the train i at tk The error covariance matrix updated at the moment, that is, the error covariance matrix at the current moment, χ i (t k-1 ) is the train i at t k-1 The error covariance matrix updated at all times, I is the identity matrix.

[0109] In one embodiment of the present invention, the k The status information of train i in the heavy-load train group is expressed by the following formula:

[0110] x(t k ):=[s(t k ), v(t k ), a(t k )] T

[0111] s(t k ):=[s1(t k ), s2(t k ),…,s n (t k )] T

[0112] v(t k ):=[v1(t k ), v2(t k ),…,v n (t k )] T

[0113] a(t k ):=[a1(t k ), a2(t k ),…,a n (t k )] T

[0114] In the formula, x(t k )=[x1(t k ),x2(t k ),...,x n (t k )] T ,x(t k ) indicates that at time t k Status information of the heavy-load train group, s i (t k ) represents the train i in the heavy-load train group at t k The position at the moment, v i (t k ) represents the train i in the heavy-load train group at t kThe speed of the moment, a i (t k ) represents the train i in the heavy-load train group at t k The acceleration at the moment, n is the number of trains in the heavy-load train group.

[0115] In one embodiment of the present invention, the error covariance matrix at the current moment is updated using the following formula:

[0116]

[0117] Where, is the minimum data packet success rate allowed by the system, χ i - (t k ) is t k The estimated error covariance matrix of the state information of train i at time t is the updated error covariance matrix, χ i - (t k-1 ) is t k-1 The estimated error covariance matrix of the state information of train i at time t is the error covariance matrix at the current time, A e,i is the known state matrix of train i, M e is a matrix of known dimensions, R e is the process noise covariance matrix.

[0118] The various embodiments of the present invention are described in a progressive manner. Similar portions between the various embodiments can be referred to in conjunction with each other. Each embodiment focuses on the differences between the other embodiments. In particular, the device embodiments are described briefly because they are generally similar to the method embodiments. For relevant portions, refer to the description of the method embodiments.

[0119] Those skilled in the art will appreciate that, in one or more of the above examples, the functions described herein may be implemented using hardware, software, firmware, or any combination thereof. When implemented using software, these functions may be stored in a computer-readable medium or transmitted as one or more instructions or codes on a computer-readable medium.

[0120] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent replacements, improvements, etc. made on the basis of the technical solution of the present invention should be included in the scope of protection of the present invention.

Claims

1. A heavy-load train group operation control method based on enhanced Kalman filter estimation, characterized in that: The method comprises: When the communication between the current train and the adjacent train in the heavy-load train group is interrupted, the error covariance matrix at the current moment is obtained; Determining estimated state information of adjacent trains based on the error covariance matrix at the current moment; After controlling the speed of the current train based on the status information of the adjacent train, determining whether the delay status information of the adjacent train has been received; If so, controlling the speed of the current train based on the delay status information of the adjacent train; If not, after updating the error covariance matrix at the current moment based on the estimated state information of the adjacent train, re-execute the step of "determining the estimated state information of the adjacent train based on the error covariance matrix at the current moment".

2. The method according to claim 1, characterized in that The estimated state information of the adjacent train is determined by the following formula: x i (t k )=A e,i x i (t k-1 )+B e,i u i (t k-1 )+w e,i (t k-1 ) y i (t k )=C e,i x i (t k-1 )+v e,i (t k-1 ) Where i is the number of the adjacent train, t k is the current moment, x i (t k )=[s i (t k ),v i (t k ),a i (t k )] T For t k The status information of train i in the heavy-load train group, s i (t k ) is the train i in the heavy-load train group at t k The position at the moment, v i (t k ) is the train i in the heavy-load train group at t k The speed of the moment, a i (t k ) is the train i in the heavy-load train group at t k The acceleration at the moment, u i (t k-1 ) is t k-1 The control input of the actuator of train i at time A e,i and B e,i is the known state matrix and control matrix of train i, w e,i (t k ) is the train i at t k The process noise at the moment, R e is the process noise covariance matrix, y i (t k ) is t k The observed state information of train i at time C e,i =[1,0,0] T is the observation matrix, ν e,i (t k ) is the time during which train i runs k The measurement noise at the moment, Ne is the covariance matrix of the measurement noise, t k The estimated state information of train i before correction obtained by the enhanced Kalman filter estimation algorithm at time x, that is, the estimated state information of the adjacent train, At time t k-1 The estimated state information of train i in the heavy-load train group is The train i at t is obtained by the enhanced Kalman filter estimation method k The error covariance matrix at time , t k The Kalman gain of train i at time instant, C e,i is the observation matrix, C e,i T C e,i The transposed matrix of At time t k The estimated state information of train i in the heavy-load train group, χ i (t k ) is the train i at t k The error covariance matrix updated at the moment, that is, the error covariance matrix at the current moment, χ i (t k-1 ) is the train i at t k-1 The error covariance matrix updated at all times, I is the identity matrix.

3. The method according to claim 2, characterized in that Said in t k The status information of train i in the heavy-load train group is expressed by the following formula: x(t k ):=[s(t k ),v(t k ),a(t k )] T s(t k ):=[s1(t k ),s2(t k ),…,s n (t k )] T v(t k ):=[v1(t k ),v2(t k ),…,v n (t k )] T a(t k ):=[a1(t k ),a2(t k ),…,a n (t k )] T Where, x(t k )=[x1(t k ),x2(t k ),...,x n (t k )] T ,x(t k ) indicates that at time t k Status information of the heavy-load train group, s i (t k ) represents the train i in the heavy-load train group at t k The position at the moment, v i (t k ) represents the train i in the heavy-load train group at t k The speed of the moment, a i (t k ) represents the train i in the heavy-load train group at t k The acceleration at the moment, n is the number of trains in the heavy-load train group.

4. The method according to claim 1, wherein The error covariance matrix at the current moment is updated using the following formula: Where, is the minimum data packet success rate allowed by the system, χ i - (t k ) is t k The estimated error covariance matrix of the state information of train i at time t is the updated error covariance matrix, χ i - (t k-1 ) is t k-1 The estimated error covariance matrix of the state information of train i at time t is the error covariance matrix at the current time, A e,i is the known state matrix of train i, M e is a matrix of known dimensions, R e is the process noise covariance matrix.

5. A heavy-load train group operation control device based on enhanced Kalman filter estimation, characterized in that: The device comprises: an acquisition unit configured to acquire an error covariance matrix at a current moment when communication between a current train and an adjacent train in the heavy-load train group is interrupted; a first data processing unit configured to determine estimated state information of adjacent trains based on the error covariance matrix at the current moment; a second data processing unit configured to, after controlling the speed of the current train based on the status information of the adjacent train, determine whether the delay status information of the adjacent train has been received; a third data processing unit, configured to, if yes, control the speed of the current train based on the delay status information of the adjacent train; The fourth data processing unit is configured to, if not, update the error covariance matrix at the current moment based on the estimated state information of the adjacent train, and then re-execute the step of "determining the estimated state information of the adjacent train based on the error covariance matrix at the current moment".

6. The device according to claim 5, characterized in that The estimated state information of the adjacent train is determined by the following formula: x i (t k )=A e,i x i (t k-1 )+B e,i u i (t k-1 )+w e,i (t k-1 ) y i (t k )=C e,i x i (t k-1 )+v e,i (t k-1 ) Where i is the number of the adjacent train, t k is the current moment, x i (t k )=[s i (t k ),v i (t k ),a i (t k )] T For t k The status information of train i in the heavy-load train group, s i (t k ) is the train i in the heavy-load train group at t k The position at the moment, v i (t k ) is the train i in the heavy-load train group at t k The speed of the moment, a i (t k ) is the train i in the heavy-load train group at t k The acceleration at the moment, u i (t k-1 ) is t k-1 The control input of the actuator of train i at time A e,i and B e,i is the known state matrix and control matrix of train i, w e,i (t k ) is the train i at t k The process noise at the moment, R e is the process noise covariance matrix, y i (t k ) is t k The observed state information of train i at time C e,i =[1,0,0] T is the observation matrix, ν e,i (t k ) is the time during which train i runs k The measurement noise at the moment, Ne is the covariance matrix of the measurement noise, t k The estimated state information of train i before correction obtained by the enhanced Kalman filter estimation algorithm at time t, that is, the estimated state information of the adjacent train, At time t k-1 The estimated state information of train i in the heavy-load train group is The train i at t is obtained by the enhanced Kalman filter estimation method k The error covariance matrix at time , t k The Kalman gain of train i at time instant, C e,i is the observation matrix, C e,i T C e,i The transposed matrix of At time t k The estimated state information of train i in the heavy-load train group, χ i (t k ) is the train i at t k The error covariance matrix updated at the moment, that is, the error covariance matrix at the current moment, χ i (t k-1 ) is the train i at t k-1 The error covariance matrix updated at all times, I is the identity matrix.

7. The device according to claim 6, characterized in that Said in t k The status information of train i in the heavy-load train group is expressed by the following formula: x(t k ):=[s(t k ),v(t k ),a(t k )] T s(t k ):=[s1(t k ),s2(t k ),…,s n (t k )] T v(t k ):=[v1(t k ),v2(t k ),…,v n (t k )] T a(t k ):=[a1(t k ),a2(t k ),…,a n (t k )] T Where, x(t k )=[x1(t k ),x2(t k ),...,x n (t k )] T ,x(t k ) indicates that at time t k Status information of the heavy-load train group, s i (t k ) represents the train i in the heavy-load train group at t k The position at the moment, v i (t k ) represents the train i in the heavy-load train group at t k The speed of the moment, a i (t k ) represents the train i in the heavy-load train group at t k The acceleration at the moment, n is the number of trains in the heavy-load train group.

8. The device according to claim 5, characterized in that The error covariance matrix at the current moment is updated using the following formula: Where, is the minimum data packet success rate allowed by the system, χ i - (t k ) is t k The estimated error covariance matrix of the state information of train i at time t is the updated error covariance matrix, χ i - (t k-1 ) is t k-1 The estimated error covariance matrix of the state information of train i at time t is the error covariance matrix at the current time, A e,i is the known state matrix of train i, M e is a matrix of known dimensions, R e is the process noise covariance matrix.

9. An electronic device, characterized in that: The method comprises a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the method according to any one of claims 1 to 4 is implemented.

10. A computer-readable storage medium, characterized in that A computer program is stored thereon, and when the computer program is executed in a computer, the computer is caused to execute the method according to any one of claims 1 to 4.

Images