A predictive control method, system, device and medium for a train group

Abstract

The present application discloses a predictive control method, system, device and medium for a train group, relating to the technical field of train predictive control. The method includes: determining the coupling relationship between each train by performing topology on the train group; the train group includes a number of trains; the coupling relationship includes one or more of the mutual traction relationship and the mutual braking relationship between each train; establishing a vehicle operation model of the train with the coupling relationship force of the train as the input and the speed of the train as the output; using a distributed predictive control algorithm, based on the coupling relationship between each train in the train group and the vehicle operation model of the train, predicting and controlling the speed of each train so that the difference between the speeds of each train in the train group is less than a set threshold. The present application can make the speeds of the trains in the train group finally stabilize within a certain speed range, effectively improving the passing capacity of the train group in the section.

Description

Technical Field

[0001] The present application relates to the technical field of train predictive control, and particularly to a predictive control method, system, device and medium for a train group. Background Art

[0002] Railway transportation has always been dominant in China's transportation industry. Although the tense situation of railway freight has eased nowadays, there are still problems such as prominent supply-demand contradictions and insufficient transport capacity in some areas. To solve these problems, train groups are usually used for transportation. For example, the heavy-haul train transportation mode can be used. When using heavy-haul train groups, freight railways undertake heavy transportation tasks, especially in the transportation of bulk goods such as coal. According to statistics, the total national railway freight volume in 2022 reached 3.903 billion tons, an increase of 177 million tons compared with the previous year, a growth of 4.8%.

[0003] In the case of the mixed operation of electric locomotives and diesel locomotives commonly adopted in heavy-haul railways, from the perspective of train operation safety, it is necessary to improve the passing capacity of sections and the safe operation of locomotive groups to ensure the smooth operation of trains and reduce the labor intensity of drivers.

[0004] Train predictive control technology can improve the operation efficiency and safety of trains. However, all existing predictive controls are for single trains and do not consider the coupling relationship between trains, which may lead to large speed control errors during train operation, limiting their application potential in more complex and multi-vehicle operation environments. Summary of the Invention

[0005] The purpose of the present application is to provide a predictive control method, system, device and medium for a train group, which can make the speeds of trains in the train group finally stabilize within a certain speed range, effectively improving the passing capacity of sections of the train group.

[0006] To achieve the above purpose, the present application provides the following solutions:

[0007] In the first aspect, the present application provides a predictive control method for a train group, including:

[0008] Determine the coupling relationship between trains by topology for the train group; the train group includes several trains; the coupling relationship includes one or more of the mutual traction relationship and the mutual braking relationship between trains;

[0009] Establish a vehicle operation model of the train with the coupling relationship force of the train as the input and the speed of the train as the output; the coupling relationship force is the traction force or the braking force;

[0010] Using a distributed predictive control algorithm, based on the coupling relationship between trains in the train group and the vehicle operation model of the trains, predictively control the speeds of the trains so that the difference between the speeds of the trains in the train group is less than a set threshold.

[0011] In a second aspect, the present application provides a predictive control system for a train group, including:

[0012] A coupling relationship determination module, configured to: perform topology on the train group to determine the coupling relationship between the trains; the train group includes a number of trains; the coupling relationship includes one or more of the mutual traction relationship and the mutual braking relationship between the trains;

[0013] A vehicle operation model establishment module, configured to: establish a vehicle operation model of the train with the coupling relationship force of the train as the input and the speed of the train as the output; the coupling relationship force is the traction force or the braking force;

[0014] A distributed predictive control module, configured to: use a distributed predictive control algorithm to predictively control the speeds of the trains based on the coupling relationship between the trains in the train group and the vehicle operation model of the trains so that the difference between the speeds of the trains in the train group is less than a set threshold.

[0015] In a third aspect, the present application provides a computer device, including: a memory, a processor, and a computer program stored on the memory and executable on the processor, where the processor executes the computer program to implement the above-mentioned predictive control method for a train group.

[0016] In a fourth aspect, the present application provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the above-mentioned predictive control method for a train group is implemented.

[0017] According to the specific embodiments provided by the present application, the following technical effects are disclosed in the present application:

[0018] The present application provides a predictive control method, system, device, and medium for a train group, performs topology on the train group to determine the coupling relationship between the trains, establishes a vehicle operation model of the train, and uses a distributed predictive control algorithm to predictively control the speeds of the trains based on the coupling relationship between the trains in the train group and the vehicle operation model of the trains so that the difference between the speeds of the trains in the train group is less than a set threshold. When performing distributed predictive control, the coupling relationship between the trains can be taken into account, so that the train group will ultimately stabilize within a certain speed range, effectively improving the section passing capacity of the train group. Description of the Drawings

[0019] To more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the accompanying drawings required for use in the embodiments. Obviously, the accompanying drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other accompanying drawings can also be obtained based on these drawings.

[0020] Figure 1 It is an application environment diagram of a predictive control method for a train group in an embodiment of the present application;

[0021] Figure 2 It is a schematic flowchart of a predictive control method for a train group provided in an embodiment of the present application;

[0022] Figure 3 It is a schematic structural diagram of an example case of the self-organizing topology model provided in an embodiment of the present application;

[0023] Figure 4 It is a schematic structural diagram of a self-organizing topology model with stable interconnection and interoperability parameters for seven locomotives provided in an embodiment of the present application;

[0024] Figure 5 It is a schematic diagram of the traction characteristic curve of the HXD3 locomotive provided in an embodiment of the present application;

[0025] Figure 6 It is a schematic diagram of the traction characteristic curve of the SS4B locomotive provided in an embodiment of the present application;

[0026] Figure 7 It is a schematic diagram of the traction characteristic curve of the HXN3 locomotive provided in an embodiment of the present application;

[0027] Figure 8 It is a schematic structural diagram of the first self-organizing topology model with stable interconnection and interoperability parameters for three locomotives provided in an embodiment of the present application;

[0028] Figure 9 It is a schematic structural diagram of the second self-organizing topology model with stable interconnection and interoperability parameters for three locomotives provided in an embodiment of the present application;

[0029] Figure 10 It is a schematic diagram of the change trend of the unit step response provided in an embodiment of the present application;

[0030] Figure 11 It is a curve graph of the control result of the distributed system of the self-organizing topology model provided in an embodiment of the present application; Figure 11 (a) is a curve graph of the speed prediction result of the distributed system of the self-organizing topology model; Figure 11 (b) is a curve graph of the traction force of the distributed system of the self-organizing topology model;

[0031] Figure 12 Schematic diagram of the functional modules of a prediction control system for a train group provided by an embodiment of the present application;

[0032] Figure 13 Schematic diagram of the structure of a computer device provided by an embodiment of the present application. Detailed implementation manners

[0033] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application.

[0034] To make the above objects, features, and advantages of the present application more obvious and understandable, the present application will be further described in detail below in conjunction with the accompanying drawings and specific implementation manners.

[0035] The prediction control method for a train group provided by the embodiments of the present application can be applied to an application environment as Figure 1 shown. Among them, the terminal 102 communicates with the server 104 through the network. The data storage system can store the data that the server 104 needs to process. The data storage system can be set up separately, integrated on the server 104, or placed on the cloud or other servers. The terminal 102 can send the train group information to the server 104. After receiving the train group information, the server 104 determines the coupling relationship between the trains by performing topology on the train group; the train group includes several trains. Taking the coupling relationship force of the trains as the input and the speed of the trains as the output, a vehicle operation model of the trains is established. Using the distributed prediction control algorithm, based on the coupling relationship between the trains in the train group and the vehicle operation model of the trains, the speeds of the trains are predicted and controlled. The server 104 can feedback the obtained speed prediction control output for each train in the train group to the terminal 102. In addition, in some embodiments, the prediction control method for the train group can also be implemented by the server 104 or the terminal 102 alone. For example, the terminal 102 can directly perform the prediction control processing on the train group, or the server 104 can perform the prediction control processing on the train group from the data storage system.

[0036] Among them, the terminal 102 can be, but is not limited to, various desktop computers, laptop computers, smart phones, and tablet computers. The server 104 can be implemented by an independent server or a server cluster composed of multiple servers, and can also be a cloud server.

[0037] In an exemplary embodiment, asFigure 2 As shown in Figure 2 , a predictive control method for a train group is provided. This method is executed by a computer device, which can be specifically executed by a computer device such as a terminal or a server alone, or jointly executed by a terminal and a server. In the embodiments of the present application, taking the application of this method to Figure 1 the server 104 in Figure 1 as an example for illustration, it includes the following steps 201 to 203. Wherein:

[0038] Step 201: Determine the coupling relationship between trains by topology for the train group; the train group includes several trains; the coupling relationship includes one or more of the mutual traction relationship and the mutual braking relationship between trains.

[0039] Step 202: Establish a vehicle operation model of the train with the coupling relationship force of the train as the input and the train speed as the output; the coupling relationship force is the traction force or the braking force.

[0040] Step 203: Use the distributed predictive control algorithm to predict and control the speeds of the trains based on the coupling relationship between the trains in the train group and the vehicle operation model of the train, so that the difference between the speeds of the trains in the train group is less than the set threshold.

[0041] Implementing the above steps 201 to 203, performing topological analysis on the train group to obtain a self-organizing topological model with a stable structure (i.e., the coupling relationship between the trains in the train group), and performing distributed predictive control on the train group through this self-organizing topological model with a stable structure. Thus, when performing distributed predictive control, the coupling relationship between trains can be taken into account, and it conforms to the self-organizing topological model. After performing distributed predictive control, the speeds of the trains in the train group will ultimately stabilize within a certain speed range, effectively improving the passing capacity of the train group in the section. In addition, the present application can also dynamically adjust the parameters in the predictive control through preset conditions. During operation, according to the actual control effect, the parameters are adjusted accordingly to make the control of the train operation reach the best effect.

[0042] In some specific embodiments, the present application is based on distributed dynamic matrix control for distributed control in the case of mixed operation of electric trains and diesel trains commonly used in busy heavy-haul railways, that is, the train group in the present application includes several electric trains and diesel trains. The present application uses a heterogeneous group modeling method to identify train groups with different dynamic characteristics, and then analyzes the dynamic coupling characteristics between different train groups to establish a self-organizing topological model of the train group. Combining the established self-organizing topological model and distributed dynamic matrix control to predict and control the train speed, preventing large speed control errors from occurring during the train operation and achieving safe and stable operation.

[0043] In step 201, taking the traction force of the train as an example, in this embodiment, a self-organizing topology model with a stable structure is established for the traction characteristics of each different train in the train group. The coupling relationship between each train forms the self-organizing topology model, and the coupling relationship includes one or more of the mutual traction relationship and the mutual braking relationship between the trains.

[0044] Based on the self-organizing topology model, distributed predictive control is performed on the train group. Information is exchanged between the trains to keep the train group satisfy the coupling relationship, so that the train group can finally stabilize within a certain speed range, which can effectively improve the passing capacity of the train group in the section and realize the safe operation of the train group.

[0045] In an exemplary embodiment, for the dynamic coupling characteristics between different trains, taking the traction force of the train as the topological object to obtain the coupling relationship between the trains, the general structure of the self-organizing topology model of the heterogeneous group as shown in Figure 3 can be obtained. In the self-organizing topology model, vehicle 1 is used as the leading vehicle for traction and needs to have a greater traction power to provide for the heavy-haul train group. The traction power of vehicles 2 to 7 decreases in turn or is similar, so as to obtain a topological graph of the heterogeneous group with stable interconnection parameters, that is, the heavy-haul train heterogeneous group model. The double-headed arrows in the heavy-haul train heterogeneous group model represent the mutual communication between the trains. The interconnection parameters refer to the vehicle characteristic parameters of different types of trains (such as traction / braking dynamics parameters, wheel-rail adhesion coefficient).

[0046] It should be understood that the topological results of different situations correspond to different self-organizing topology models. Therefore, in different embodiments, due to different train groups, the corresponding self-organizing topology model should be obtained to achieve an ideal control effect.

[0047] Combined with the foregoing content, a self-organizing topology model of seven trains can be obtained, as specifically shown in Figure 4 as follows.

[0048] In another exemplary embodiment of the present application, the above step 202 can be replaced by the following steps 301 to 302:

[0049] Step 301: Taking the traction force of each train as the input and the speed of each train as the output, establish a target transfer function; the target transfer function is used to describe the step response of the train;

[0050] Step 302: Describing the traction process of the train with the discretized structure of the target transfer function, and establishing a vehicle operation model of the train.

[0051] Among them, step 301 specifically includes the following steps 401 to 402:

[0052] Step 401: Identify the object to be identified to obtain identification parameters;

[0053] Step 402: Use the traction force of each train as the input and the speed of each train as the output, and establish a target transfer function using a first-order model with time delay.

[0054] In an exemplary embodiment, a vehicle operation model with time delay can be established based on the train traction characteristic curve.

[0055] In this embodiment, according to the traction characteristics, the target transfer function is established through the step response of the controlled object. Then, a first-order model with time delay is used to approximately describe the vehicle step response. The relationship between the transfer function F(s) of the train traction force and the transfer function v(s) of the train speed is the target transfer function G(s). The expression of the target transfer function is as follows:

[0056]

[0057] Where, G(s) is the target transfer function; F(s) is the transfer function of the train traction force; v(s) is the transfer function of the train speed; Y(s) is the speed prediction output vector; U(s) is the control input vector; τ is the time delay constant; s represents the complex variable; a and b are identification parameters.

[0058] Step 401 specifically includes: using the least squares algorithm to identify the object to be identified to obtain identification parameters.

[0059] Combined with the above formula, it can be seen that the target transfer function depends on parameters a and b. In different embodiments, parameters a and b can be obtained in various ways. In an exemplary embodiment, the method of matrix control can be used to implement predictive control, and parameters a and b are identified based on the least squares algorithm. In different embodiments, different methods can also be used to obtain the foregoing parameters.

[0060] In an exemplary embodiment, obtain the sampling values of the step response:

[0061] a j = a(jT), j = 1, 2, …, N (2);

[0062] Where, T is the sampling period, N is the modeling time domain, and the set of a j constitutes the model vector of the step response.

[0063] Obtain the vehicle operation model of the train by describing the train traction process with the discretized structure of the target transfer function. Thus, the discretized structure of the target transfer function can be described by a first-order autoregressive model with time delay to describe the train traction process. Then, the vehicle operation model of the train is as follows:

[0064] y(k) = -a1y(k - 1 - τ) + b0u(k - 1) + ξ(k) (3);

[0065] Among them, y(k) represents the train speed at time k, y(k - 1 - τ) represents the train speed at time k - 1 - τ, τ represents the time delay constant, u(k - 1) represents the traction force of the train at time k - 1, a1 represents the first parameter, b0 represents the second parameter, and ξ(k) represents the noise sequence at time k.

[0066] In order to better carry out the subsequent process, in an exemplary embodiment, the vehicle operation model of the train is also expressed as:

[0067]

[0068] Among them, is the data vector, and θ = [a1 b0] represents the parameter to be identified.

[0069] Then the identification object is the parameter a, b or the first parameter a1, the second parameter b0. The first parameter a1 and the second parameter b0 respectively correspond to the identification parameters a and b in the target transfer function. In the foregoing process, the identification parameters are obtained by the least squares algorithm, and then the first parameter a1 and the second parameter b0 of each train are obtained.

[0070] In an exemplary embodiment, based on the least squares algorithm, the first parameter and the second parameter are identified using the actual parameters, and the final result obtained by the identification is used as the first parameter and the second parameter of the vehicle operation model to obtain the vehicle operation model. It should be noted that in different embodiments, the identification process may also be different. For example, different algorithm models are used to implement the identification process.

[0071] In an exemplary embodiment, a distributed dynamic matrix self-organizing network system control can be adopted. According to the dynamic response signal, predictive control is performed on the speed output by the vehicle operation model, where the dynamic response signal is a step response signal.

[0072] In this embodiment, the distributed predictive control algorithm can establish communication between several trains. Each train can obtain the state information of other trains and incorporate it into the prediction model for prediction. Therefore, the coupling relationship between systems can be considered. Applying the control method of this embodiment in practical engineering applications can consider the interaction information between various parts in the train group, and the association effect and causal relationship between non-adjacent trains.

[0073] As described above, the distributed dynamic matrix predictive control method can consider the association effect between all trains.

[0074] The communication methods between trains can be various, including but not limited to wireless network communication, satellite communication, and communication based on the ground signal system. For example, by using wireless communication technologies such as LTE or 5G networks, trains can exchange key information in real time, such as speed, position, and estimated arrival time. Or, through satellite communication, the signals between trains are transmitted via satellites, enabling trains to maintain a stable communication connection even in remote areas. Communication based on the ground signal system can also be used, etc.

[0075] In different situations, trains exchange information with each other. The coupling relationship among the trains in the train group includes the relationship of mutual influence and interaction among multiple trains during operation. This coupling relationship may be caused by factors such as the mutual traction relationship and mutual braking relationship among trains, etc.

[0076] In an exemplary embodiment, the coupling relationship in the train group is complex and variable. For example, when a train decelerates due to a fault, the train following it needs to adjust its speed in a timely manner to avoid collision. At the same time, the train ahead may also need to adjust its speed to maintain the smooth operation of the entire train group. An exemplary embodiment adopts a self-organizing topology model. Through this model, the train group automatically organizes into a stable topological structure. Each train acts not only as an independent control unit but also as a node in the entire group. Through communication and coordination with surrounding trains, the optimal operation of the group is jointly achieved. The coordination among trains can be realized through any of the aforementioned communication methods. Of course, a master control unit can also be set to conduct overall control over each train, etc.

[0077] In this embodiment, the control strategy of the train group is based on the distributed predictive control algorithm. The distributed predictive control algorithm allows each train to independently calculate its optimal control strategy according to its own state and the information collected from other trains. This strategy not only considers the operation efficiency of a single train but also the overall performance of the entire train group, thereby significantly improving the passing capacity of the train group in the section while ensuring safety.

[0078] In an exemplary embodiment, the distributed predictive control algorithm is the distributed dynamic matrix predictive control algorithm. The above step 203 specifically includes: The speed change process of the train group is composed of n subsystems. The prediction equation of the i-th subsystem is:

[0079] y i (k) = f i [y i,0 (k), Δu1(k), …, Δu n (k)] (5);

[0080] Among them, y i(k) represents the predicted output value of the i-th subsystem at a future time, f is a mapping vector function, y i,0 (k) represents the initial predicted value of the i-th subsystem at time k, Δu i (k) represents the control input vector of the i-th subsystem at time k, where i = 1, 2, …, n.

[0081] The distributed dynamic matrix control adopted in this application is an optimal control technology. It can effectively solve the time-delay process problem and implement control according to the quadratic performance index with the smallest deviation between the predicted output and the given value. The minimum quadratic performance index is the performance index of the distributed dynamic matrix control algorithm. According to the proportional and superposition characteristics of the linear system, the output of the system at a future time can be predicted based on the given input.

[0082] Among them, the minimum quadratic performance index of the distributed dynamic matrix control algorithm can be expressed as:

[0083]

[0084] Among them, is the minimum quadratic performance index, represents the i-th subsystem, and the optimal control sequence at the (l + 1)-th iteration, where i = 1, 2, …, n; M is the control time domain; j = 1, 2, …, P, P is the prediction time domain, the error weighting matrix Q is the P×P identity matrix, the control weighting matrix R is the M×M identity matrix, Y i,M (k) represents the predicted output value of the system, and W(k) represents the set value.

[0085] In a specific case, the distributed dynamic matrix control algorithm is specifically as follows: Among the n subsystems, the output of the entire system predicted within the next P time instants is:

[0086] Y(k + j) = f[Y0(k), Δu1(k), …, Δu n (k)] (7);

[0087] Among them, time j = 1, 2, …, P; Δu i represents the M i -dimensional control input vector; Y(k + j) and Y0(k) respectively represent the predicted output value of the system at a future time and the initial predicted value at time k.

[0088] The predicted output model of the system can be expressed as:

[0089] Y M (k + 1) = Y0(k + 1) + AΔU(k) (8);

[0090] Among them, Y M (k + 1) = [y M(k + 1)y M (k + 2)…y M (k + P)] T , Y0(k + 1) represents the initial prediction value at time k + 1 of the system, and ΔU(k) = [Δu(k) Δu(k + 1) … Δu(k - M + 1)] T ; A represents the dynamic matrix. The expression of the dynamic matrix A is as follows:

[0091]

[0092] During the train operation, due to the existence of external disturbances, the predicted result may deviate from the predicted value. Therefore, the output prediction value of the system needs to be corrected by weighting with the actual output error on the basis of the prediction model. The correction expression is as follows:

[0093] Y cor (k + 1) = Y m (k + 1) + h[Y(k + 1) - Y m (k + 1)] = AΔU(k) + A0ΔU(k) + he(k) (10);

[0094] Among them, Y cor (k + 1) = [y cor (k + 1)y cor (k + 2)…y cor (k + P)] T represents the corrected output prediction vector, h = [h1 h2…h P T represents the feedback correction vector, including the feedback correction values at the next P time instants; Y m (k + 1) represents the speed prediction value during train operation; e(k) represents the output error, that is, the difference between the actual output of the train speed at time k and the calculated speed prediction value.

[0095] The corrected prediction vector Y cor (k + 1) can be used to deduce the initial prediction value at time k + 1 by shifting the time base from time k to time k + 1 through the shift matrix S:

[0096] Y N (k + 1) = SY cor (k + 1) (11);

[0097] Among them, S represents the shift matrix, and the expression of the shift matrix S is as follows:

[0098]

[0099] ​The distributed predictive control algorithm in this application is not limited to the distributed dynamic matrix control algorithm, and other distributed predictive control algorithms can also be applicable.

[0100] In another exemplary embodiment of this application, the above step 203 further includes: dynamically adjusting the online operation parameters according to preset adjustment conditions; the online operation parameters include sampling period, model time domain, prediction time domain, error weight matrix, control time domain, control right matrix, and correction parameters.

[0101] In different implementation cases, during online operation, it is necessary to adjust the parameters accordingly according to the actual control effect. Therefore, in some exemplary embodiments, the distributed predictive control process further includes a dynamic parameter adjustment process. The dynamic parameter adjustment specifically dynamically adjusts the parameters according to the preset adjustment conditions and obtains the corresponding parameters according to the preset conditions.

[0102] For example, in the case of adopting distributed matrix predictive control:

[0103] (1) Sampling period T and model time domain N:

[0104] Dynamically adjust the sampling period T and the model time domain N. The preset adjustment condition is that the sampling period T needs to satisfy the Shannon sampling theorem, and at the same time, the selection of T should follow the principle of T selection in general sampling control systems.

[0105] The foregoing conditions are determined according to the type of the controlled object and its dynamic characteristics. Due to the differences in the type of the controlled object and its dynamic characteristics, the range of the sampling period T is from the millisecond order of magnitude to the order of magnitude of dozens of seconds.

[0106] The selection of the sampling period T is related to the model time domain N. Usually, the model time domain N is set within the range of 20 to 50. Therefore, if T is selected too small, it will increase the calculation frequency and also lead to an increase in the amount of calculation. However, from the perspective of anti-interference, it is hoped that the system can quickly and timely suppress the influence of interference.

[0107] (2) Similarly, in the case of distributed matrix predictive control, the preset conditions for the prediction time domain P and the error weight matrix Q are: when the system makes predictions, the prediction time domain P should include all responses that have a greater impact on the current moment control. For a system with time delay, the optimized time domain P should exceed the time delay part of the object's step response. Therefore, the weight coefficient corresponding to the time delay part can be taken as 0.

[0108] The size of the prediction horizon \(P\) has a significant impact on the rapidity and robustness of the system. If \(P\) is taken to be small enough, such as \(P = 1\), rapid one-step optimal control can be achieved. However, there will be ripples between sampling points, and at the same time, the robustness to model mismatch and disturbances is poor. Therefore, it is not applicable to systems with time delay. Thus, \(P\) in the distributed system control of heavy-haul train groups should not be taken too small. If \(P\) is taken to be quite large, it can enhance the stability of the system and can derive a stable control law. However, it will have an impact on the rapidity of the system. Therefore, \(P\) in the preset condition control system cannot be taken too large.

[0109] (3) Control horizon \(M\) and control weight matrix \(R\):

[0110] For another example, the preset condition for the control horizon \(M\) and the control weight matrix \(R\) is \(M\leq P\).

[0111] During execution, if \(M\) is taken to be very large, approaching \(P\), then the maneuverability of the control will be enhanced and the dynamic response of the system will be improved, but the robustness and stability of the system will become poor. Therefore, \(M\) cannot be taken too large. If \(M\) is taken to be very small, such as \(M = 1\), although the robustness and stability of the system are enhanced, the maneuverability of the system control is weak, and it is difficult to ensure that the output can closely track the expected value between sampling points. Therefore, the preset condition is to first select \(M\) according to the dynamic characteristics of the controlled object in the distributed predictive control system of heavy-haul trains, and then adjust \(P\).

[0112] The control weight matrix \(R\) is also taken as a diagonal matrix in the system, that is, \(R=\text{diag}\{r_1,r_2,\cdots,r_n\}\). Among them, usually \(r_i\) is taken as the same value. The main purpose of adjusting \(r_i\) is to prevent the control quantity from changing too violently. The role of the control weight matrix \(R\) is to moderately limit the violent change of the control increment. Therefore, \(r\) can be set to \(0\) first. If the corresponding control system tends to be stable but the control quantity changes violently, then \(r\) should be appropriately increased at this time. For the distributed predictive control system of heavy-haul trains, in order to make the control system stable and at the same time reduce the change of the control quantity, only a very small value needs to be taken for \(r\). M} Among them, usually \(r_i\) j is taken as the same value. For \(r_i\) j The main purpose of adjustment is to prevent the control quantity from changing too violently. The role of the control weight matrix \(R\) is to moderately limit the violent change of the control increment. Therefore, \(r\) can be set to \(0\) first. If the corresponding control system tends to be stable but the control quantity changes violently, then \(r\) should be appropriately increased at this time. For the distributed predictive control system of heavy-haul trains, in order to make the control system stable and at the same time reduce the change of the control quantity, only a very small value needs to be taken for \(r\).

[0113] (4) Correction coefficient:

[0114] For another example, the correction coefficient \(h\) i , the coefficient \(h\) i will not affect the dynamic response of the control. It will play a role when there is a model mismatch in the object or it is subject to unknown disturbances. Its selection is not affected by other parameters. For the distributed predictive control system of heavy-haul trains, \(h_1 = 1\) can be taken, \(h\) i= α, i = 2, …, N, 0 < α ≤ 1. When α decreases, the robustness of the control system will be enhanced. When α approaches 0, the system has strong robustness but poor anti-interference ability. When α approaches 1, the anti-interference ability of the system is enhanced, but the robustness is weakened. Therefore, when determining the value of α, the robustness and anti-interference ability of the system should be considered.

[0115] There can be various specific implementation methods for the foregoing preset conditions. For example, they can be implemented through different logical judgment processes and other means.

[0116] In this embodiment, for the predictive control problem of a heavy-haul train group with time delay, by analyzing the dynamic coupling characteristics between different trains, a topological graph with stable interconnection parameters for the heterogeneous group can be obtained. Using the traction force as the control signal as the control input and the train speed as the output, a vehicle operation model is established. Based on distributed system control, predictive control is performed on the train speed to improve the passing capacity of the section and achieve safe operation of the train group.

[0117] In another exemplary embodiment of the present application, taking three trains as an example for simulation verification, the self-organizing topology model mainly considers the mutual traction relationship between each train. The train group includes three types of trains: the AC electric train HXD3, the DC electric train SS4B, and the diesel train HXN3. The traction characteristic curves of the three trains are as Figures 5 to 7 shown, and the foregoing three trains form a train group.

[0118] Among them, HXD3 is a single-phase industrial frequency AC electric train using an AC-DC-AC drive mode. Compared with the SS4B DC electric train using a DC drive mode, it has better traction performance. The HXN3 diesel AC train has relatively poor traction performance compared with the previous two types of trains.

[0119] In this embodiment, considering the traction force, the self-organizing topology models of the above three trains HXD3, SS4B, and HXN3 are obtained through topology. The coupling relationships between the above three trains HXD3, SS4B, and HXN3 are as Figure 8 and Figure 9 shown. The target transfer function of the controlled object (the above three trains HXD3, SS4B, and HXN3) consists of a time-delay link and an inertia link. The change trend of the step response is as Figure 10 .

[0120] In this embodiment, due to the traction characteristics of the train, the mathematical model of the controlled object is uncertain. Therefore, when the train levels are different, the mathematical models of the controlled objects are also different. Therefore, in this embodiment, the least squares method is used to identify the target transfer function of each train respectively, and the vehicle operation models of each train in the train group can be determined.

[0121] In this embodiment, the traction data at the 6th level of the HXD3 train, the 2nd level of the train handle of the SS4B train, and the 1st level of the HXN3 train are selected respectively. Since the dynamic response of DC trains is much slower than that of AC trains, in order to ensure the stability of system control, the lag time of the SS4B DC train needs to be greater than that of the HXD3 and HXN3 trains. The target transfer function is obtained as follows:

[0122]

[0123] Among them, the target transfer function G ij is determined as follows:

[0124]

[0125] In the formula, i = 1, 2, 3; j = 1, 2, 3.

[0126] The target transfer functions of the three trains are respectively:

[0127]

[0128] In this embodiment, to ensure the smooth operation of the train group and prevent large speed control errors from occurring during the train operation, the speed of heavy-haul trains is predicted and controlled through the distributed dynamic matrix control algorithm.

[0129] In this embodiment, the dynamic parameters of various online operation parameters are adjusted, and the following parameters are obtained according to the preset conditions:

[0130] The sampling period T = 1s, the modeling time domain N = 70, the optimization time domain P = 8, the control time domain M = 3, h = [1,..., 1] 1×N , the simulation time is 800s, the expected output set values of each system are (80, 75, 70), the error weighting matrix Q is the unit matrix of P×P, the control weighting matrix R is the unit matrix of M×M, and the error weight matrix is set as Q = diag{1, 1,..., 1}.

[0131] The control right matrix is set as R1 = diag{0.22, 0.22,..., 0.22}, R2 = diag{0.5, 0.5,..., 0.5}, R3 = diag{2.5, 2.5,..., 2.5}.

[0132] The foregoing parameters and the first parameter and the second parameter obtained by identification are used for distributed matrix predictive control, and the results are referred to Figure 11 . From Figure 11 the speed prediction results, it can be seen that the speed prediction control effect of this embodiment matches the speed set value, and the control performance is good.

[0133] The present application also provides an application scenario, which applies the above-mentioned predictive control method for a train group. Specifically: The predictive control method for a train group provided in this embodiment can be applied to the predictive control scenario of a train group. The predictive control scenario of a train group includes a train group information acquisition link, a predictive control link, and an operation link; the train group enters the predictive control link from the train group information acquisition link, obtains the predictive control output of the speed of each corresponding train, and enters the downstream operation link. The predictive control method for a train group provided in this embodiment belongs to the predictive control link. Specifically, in the process of the predictive control link for a train group, the coupling relationship between trains can be determined through topology; the train group includes several trains. Taking the coupling relationship force of the trains as the input and the speed of the trains as the output, a vehicle operation model of the trains is established. Using the distributed predictive control algorithm, based on the coupling relationship between the trains in the train group and the vehicle operation model of the trains, the speed of each train is predictively controlled.

[0134] Based on the same inventive concept, the embodiment of the present application also provides a predictive control system for a train group for implementing the above-mentioned predictive control method for a train group. The implementation solutions provided by this system to solve problems are similar to the implementation solutions described in the above method. Therefore, the specific limitations in one or more embodiments of the predictive control system for a train group provided below can refer to the limitations on the predictive control method for a train group in the above text, and will not be repeated here.

[0135] In an exemplary embodiment, as Figure 12 shown, a predictive control system for a train group is provided, including:

[0136] A coupling relationship determination module T1, configured to: determine the coupling relationship between trains through topology for the train group; the train group includes several trains; the coupling relationship includes one or more of the mutual traction relationship and the mutual braking relationship between trains;

[0137] A vehicle operation model establishment module T2, configured to: establish a vehicle operation model of the trains with the coupling relationship force of the trains as the input and the speed of the trains as the output; the coupling relationship force is the traction force or the braking force;

[0138] A distributed predictive control module T3, configured to: use the distributed predictive control algorithm to predictively control the speed of each train based on the coupling relationship between the trains in the train group and the vehicle operation model of the trains, so that the difference between the speeds of the trains in the train group is less than the set threshold.

[0139] In an exemplary embodiment, a computer device is provided. This computer device can be a server or a terminal, and its internal structure diagram can be as Figure 13As shown in the figure. The computer device includes a processor, a memory, an input / output interface (Input / Output, abbreviated as I / O), and a communication interface. Among them, the processor, the memory, and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store the predictive control data of the train group. The input / output interface of the computer device is used to exchange information between the processor and external devices. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, it implements a predictive control method for a train group.

[0140] Those skilled in the art can understand that Figure 13 the structure shown in the figure is only a block diagram of some structures related to the solution of this application, and does not constitute a limitation on the computer device to which the solution of this application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine some components, or have a different component layout.

[0141] In an exemplary embodiment, a computer device is further provided, including a memory and a processor. A computer program is stored in the memory, and when the processor executes the computer program, the steps in the above method embodiments are implemented.

[0142] In an exemplary embodiment, a computer-readable storage medium is provided, storing a computer program, and when the computer program is executed by the processor, the steps in the above method embodiments are implemented.

[0143] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use, and processing of relevant data need to comply with relevant regulations.

[0144] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to a memory, database, or other medium used in the various embodiments provided in this application can include at least one of non-volatile and volatile memories. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc.

[0145] The databases involved in the various embodiments provided in this application can include at least one of relational databases and non-relational databases. Non-relational databases can include distributed databases based on blockchain, etc., without limitation. The processors involved in the various embodiments provided in this application can be general-purpose processors, central processors, graphics processors, digital signal processors, programmable logic devices, data processing logics based on quantum computing, etc., without limitation.

[0146] The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity of description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as the scope described in this specification.

[0147] Specific examples are used in this article to elaborate on the principles and implementation manners of this application. The descriptions of the above embodiments are only used to help understand the method and its core idea of this application; at the same time, for those of ordinary skill in the art, according to the idea of this application, there will be changes in the specific implementation manners and application scopes. In summary, the content of this specification should not be construed as a limitation to this application.

Claims

1. A predictive control method for a train group, characterized in that, The predictive control method for the train group includes: Determining the topology of the train group to identify the coupling relationships between the trains; the train group includes a number of trains; the coupling relationships include one or more of the mutual traction relationships and mutual braking relationships between the trains; performing topological analysis on the train group to obtain a self-organizing topological model with a stable structure, and the self-organizing topological model is the coupling relationship between the trains in the train group; Taking the coupling relationship force of the train as the input and the train speed as the output, establishing a vehicle operation model of the train; the coupling relationship force is the traction force or the braking force; Using a distributed predictive control algorithm, based on the coupling relationships between the trains in the train group and the vehicle operation model of the train, predicting and controlling the speeds of the trains so that the difference between the speeds of the trains in the train group is less than a set threshold; the distributed predictive control algorithm is a distributed dynamic matrix predictive control algorithm, and the speed change process of the train group is composed of n subsystems, where the prediction equation of the i-th subsystem is: y i (k) = f i [y i,0 (k), Δu1(k), …, Δu n (k)]; where y i (k) represents the predicted output value of the i-th subsystem at a future time, f is a mapping vector function, y i,0 (k) represents the initial predicted value of the i-th subsystem at time k, Δu i (k) represents the control input vector of the i-th subsystem at time k, i = 1, 2, …, n; The predictive output model of the system can be expressed as: Y M (k + 1)=Y0(k + 1)+AΔU(k); Among them, Y M (k + 1) = [y M (k + 1)y M (k + 2)…y M (k + P)] T , Y0(k + 1) represents the initial prediction value at time k + 1 of the system, ΔU(k) = [Δu(k) Δu(k + 1) … Δu(k - M + 1)] T ; A represents the dynamic matrix; The predicted output value of the system needs to be corrected by the weighted method of the actual output error based on the prediction model; Dynamically adjusting the online operation parameters according to preset adjustment conditions; the online operation parameters include the sampling period, the model time domain, the prediction time domain, the error weight matrix, the control time domain, and the control right matrix.

2. The predictive control method for a train group according to claim 1, wherein Taking the coupling relationship force of the train as the input and the train speed as the output, establishing a vehicle operation model of the train, specifically including: Taking the traction force of each train as the input and the speed of each train as the output, establishing a target transfer function; the target transfer function is used to describe the step response of the train; Describing the traction process of the train with the discretized structure of the target transfer function, and establishing a vehicle operation model of the train.

3. The predictive control method for a train group according to claim 2, characterized in that, Taking the traction force of each train as the input and the speed of each train as the output, establishing a target transfer function, specifically including: Identifying the identification object to obtain identification parameters; Taking the traction force of each train as the input and the speed of each train as the output, establishing a target transfer function using a first-order model with time delay, and the expression of the target transfer function is as follows: Where, G(s) is the target transfer function; F(s) is the transfer function of the train traction force; v(s) is the transfer function of the train speed; Y(s) is the speed prediction output vector; U(s) is the control input vector; τ is the time delay constant; s represents the complex variable; a, b are identification parameters.

4. The predictive control method for a train group according to claim 3, characterized in that Identifying the identification object to obtain identification parameters, specifically including: Using the least squares algorithm to identify the identification object to obtain identification parameters.

5. The predictive control method for a train group according to claim 1, characterized in that, The vehicle operation model of the train is as follows: y(k) = -a1y(k - 1 - τ) + b0u(k - 1) + ξ(k); Where, y(k) represents the train speed at time k, y(k - 1 - τ) represents the train speed at time k - 1 - τ, τ represents the time delay constant, u(k - 1) represents the train traction force at time k - 1, a1 represents the first parameter, b0 represents the second parameter, and ξ(k) represents the noise sequence at time k.

6. A predictive control system for a train group, characterized in that, The predictive control system for the train group includes: A coupling relationship determination module, configured to: determine the coupling relationships among trains through topology analysis of a train group; the train group includes a plurality of trains; the coupling relationships include one or more of the mutual traction relationships and mutual braking relationships among the trains. A vehicle operation model establishment module, configured to: establish a vehicle operation model of a train with the coupling relationship force of the train as the input and the speed of the train as the output; the coupling relationship force is the traction force or the braking force; perform topology analysis on the train group to obtain a self-organizing topology model with a stable structure, and the self-organizing topology model is the coupling relationship among the trains in the train group. A distributed predictive control module, configured to: use a distributed predictive control algorithm to perform predictive control on the speeds of the trains based on the coupling relationships among the trains in the train group and the vehicle operation model of the train, so that the difference between the speeds of the trains in the train group is less than a set threshold; the distributed predictive control algorithm is a distributed dynamic matrix predictive control algorithm, and the speed change process of the train group is composed of n subsystems, and the prediction equation of the i-th subsystem is: y i (k) = f i [y i,0 (k), Δu1(k), …, Δu n (k)]; where y i (k) represents the predicted output value of the i-th subsystem at a future time, f is a mapping vector function, y i,0 (k) represents the initial predicted value of the i-th subsystem at time k, Δu i (k) represents the control input vector of the i-th subsystem at time k, i = 1, 2, …, n; The predictive output model of the system can be expressed as: Y M (k + 1)=Y0(k + 1)+AΔU(k); Among them, Y M (k + 1) = [y M (k + 1)y M (k + 2)…y M (k + P)] T , Y0(k + 1) represents the initial predicted value at time k + 1 of the system, ΔU(k) = [Δu(k) Δu(k + 1) … Δu(k - M + 1)] T ; A represents the dynamic matrix; The predicted output value of the system needs to be corrected by weighting with the actual output error based on the prediction model. Dynamically adjust the online operation parameters according to preset adjustment conditions; the online operation parameters include the sampling period, the model time domain, the prediction time domain, the error weight matrix, the control time domain, and the control right matrix.

7. A computer device, comprising: A memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the processor executes the computer program to implement the predictive control method for a train group according to any one of claims 1-5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the predictive control method for a train group according to any one of claims 1-5.

Images