Freight train dispatching and control system and method

Abstract

The present invention discloses a dispatching and control system and method for freight trains, which comprises the following steps: S1, an information interface device acquires basic line data and environmental information, and sends the information to a train operation status prediction system; the dispatching system interacts with the train control system to acquire train operation data in real time, and sends the information to the train operation status prediction system; S2, a train characteristic identification module establishes a train operation control model; S3, a train operation planning module predicts the train operation status based on an optimization strategy, generates a train operation time and a train control curve, and sends the results to the dispatching system; S4, the dispatching system determines whether the train operation time is reasonable, adjusts the operation plan according to the reasonable train operation time, and sends the train control curve to the train control system; the present invention can accurately predict the train operation status, thereby realizing precise adjustment of the train operation plan and refinement of train control.

Description

Technical Field

[0001] The present invention relates to the field of freight train operation, and in particular to a freight train dispatching and control system and method. Background Art

[0002] Freight railway transport often fails to complete its transportation tasks according to the established plan due to factors such as the route, environment, vehicle performance, loading and unloading, and oncoming vehicles. In such cases, the freight train operation plan needs to be adjusted to adjust the operation of the entire line. Currently, when a train fails to run as planned, the dispatcher mainly relies on personal experience to make adjustments. Since the dispatcher cannot fully predict the train operation status based on multiple factors such as the route, environment, and train performance, the adjustment accuracy is greatly deviated, which has a significant impact on the transportation efficiency of the entire line. In addition, the response speed and timeliness of manual adjustments are also delayed.

[0003] Therefore, how to accurately predict train operation status and provide a basis for precise adjustment of scheduling plans and accurate train operation has become a key issue in the integration of freight scheduling and control. At present, the research on integrated scheduling and control mainly focuses on the identification of train operation status for high-speed rail passenger trains, and does not consider the impact of factors such as line and vehicle characteristics on train operation status in freight railways. In addition, the analysis of train operation status in high-speed rail and urban rail transit is mainly based on automatic driving. Although it can achieve relatively accurate predictions, due to the complexity of high-speed rail and urban rail train performance and line conditions, they are not suitable for freight trains. Therefore, the existing methods of train operation status analysis based on automatic driving are not suitable for freight train operation status analysis.

[0004] There are also some studies on the operation status analysis of freight trains. The research direction is mainly biased towards fuzzy prediction, that is, predicting the subsequent operation status through analysis and statistics of the operation status of freight trains before the current time. This method only considers the previous operation of the train and can only calculate the subsequent train operation status when there are no abnormal conditions. However, when some special circumstances occur, such as train performance degradation, line abnormalities, bad weather and other unexpected conditions, this method cannot calculate the available train operation status, and it is impossible to achieve accurate scheduling adjustments and refined train control.

[0005] The statements herein merely provide background information related to the present invention and do not necessarily constitute prior art. Summary of the Invention

[0006] The purpose of the present invention is to provide a freight train dispatching and control system and method, which can accurately predict the running situation ahead of the freight train and improve the running efficiency.

[0007] To achieve the above object, the present invention proposes a method for dispatching and controlling freight trains, comprising the following steps:

[0008] S1. The information interface device obtains basic line data and environmental information of the line where the train is located in real time and sends it to the train operation status prediction system; the dispatching system and the train control system exchange data to obtain train operation data in real time and send it to the train operation status prediction system;

[0009] S2. The train characteristic identification module in the train operation status prediction system establishes a train operation control model based on the LSTM neural network algorithm according to the basic line data, environmental information and the train operation data. The train operation control model includes parameters of the train traction and braking systems;

[0010] S3. The train operation planning module in the train operation situation prediction system predicts the train operation situation based on the train operation control model, the basic line data, the environmental information and the train operation data based on the optimization strategy, generates the train operation time and train control curve, and sends it to the dispatching system;

[0011] S4. The dispatching system determines whether the train running time is reasonable. If it is reasonable, the operation plan will be adjusted according to the train running time, and the train control curve will be sent to the train control system to control the operation of the freight train.

[0012] Optionally, in step S4, when the dispatching system determines that the train running time conflicts with other train plans, it sends delay information to the train operation planning module;

[0013] The train operation planning module formulates a new train operation time and train control curve based on the train operation control model, the basic line data, environmental information, the train operation data and the delay information, and sends it to the dispatching system; repeats this process until the dispatching system determines that the train operation time is reasonable.

[0014] Optionally, step S2 specifically includes:

[0015] S21, based on the basic line data, environmental information and historical data of the train operation data, the train characteristic identification module establishes a basic train operation control model based on the LSTM neural network algorithm;

[0016] S22. Apply the train operation control basic model to the train operation status prediction system, and dynamically adjust the train operation control basic model according to the basic line data, environmental information and real-time data of the train operation data to obtain the train operation control model.

[0017] Optionally, in step S22, during the process of dynamically adjusting the train operation control basic model, the prediction result of the train operation control basic model is compared with the actual train operation data. If the difference between the two is within a threshold range, the train operation control basic model at this time is used as the train operation control model.

[0018] Optionally, after step S22, the prediction result of the train operation control model and the actual train operation data are continuously monitored. If the difference between the two exceeds the threshold range, the train operation control model is continuously dynamically adjusted.

[0019] Optionally, in step S3, the formula of the optimization strategy is

[0020]

[0021] Among them, J is the objective function, w1 is the weight of train traction energy consumption, F k is the traction force of the train per step, N is the total step length in the train section, Δs is the length of each step, w2 is the weight of the train following the target speed, which is directly related to the train section running time, E k is the kinetic energy of the train per step, m is the mass of the train in kg, v desk is the following speed of the train per step, in m / s.

[0022] Optionally, in step S3, the train operation planning module calculates the optimized objective function J based on the formula of the optimization strategy according to the real-time data of the train operation control model and the train operation data, and then obtains its corresponding parameters to calculate the corresponding train operation time and train control curve.

[0023] In step S3, the optimized objective function J is calculated under the equality constraints of the dynamic equation and the upper and lower limit constraints of the traction force F, the braking force B, and the kinetic energy E;

[0024] The equality constraint of the kinetic equation is E k =E k-1 +F k Δs-B k ·Δs-(W i,k +

[0025] W r,k +W a,k )·Δs, where W i,k is the slope resistance, B k is the braking force, W r,k is the curve resistance, W a,k It is the basic resistance of train operation;

[0026] The upper and lower limits of the traction force F, braking force B, and kinetic energy E are: 0≤F≤F max , 0≤B≤B(E), 0≤E≤E desk , where F max is the maximum traction force that the train can provide during operation, B(E) is the braking force that the train can provide under the current kinetic energy, and E desk It is the maximum kinetic energy of the train during operation.

[0027] Optionally, the train operation planning module divides the travel process of the freight train into a station start-up and departure phase, a station parking and braking phase, and a section operation phase. The section operation phase also includes multiple scenarios including a long downhill slope, a long uphill slope, a phase transition, a deceleration zone, and a constant speed zone.

[0028] In step S3, the train operation planning module calculates the objective function J in each stage and scenario within a certain distance ahead of the train, and then takes the sum of the objective functions as the final objective function J', calculates the parameters corresponding to the optimized final objective function J', and finally calculates the corresponding train operation time and train control curve.

[0029] Optionally, when the train operation planning module formulates a new train operation time and train control curve, if there is a constant speed zone scene within a certain distance ahead of the train, the parameters in the constant speed zone scene are adjusted first, and then the parameters in other stages and scenes are adjusted.

[0030] Optionally, the train operation data includes train control data and train operation status data, the train control data includes the operation level, current, and speed of the train control system, and the train operation status data includes the train speed and acceleration.

[0031] The present invention also proposes a freight train dispatching and control system for implementing the freight train dispatching and control method, comprising: an information interface device, a train operation status prediction system and a dispatching system;

[0032] The information interface device is used to obtain basic line data and environmental information and send it to the train operation status prediction system;

[0033] The train operation status prediction system is also connected to the dispatching system, and the dispatching system sends train operation data to the train operation status prediction system;

[0034] The train operation situation prediction system includes a train characteristics identification module and a train operation planning module;

[0035] The train characteristic identification module establishes a train operation control model based on the LSTM neural network algorithm according to the basic line data, environmental information and the train operation data;

[0036] The train operation planning module predicts the train operation status based on the train operation control model, the basic line data, the environmental information and the train operation data based on the optimization strategy, generates the train operation time and train control curve, and sends it to the dispatching system;

[0037] The dispatching system also interacts with the train control system data to obtain the train operation data in real time; the dispatching system determines whether the train operation time needs to be adjusted. If no adjustment is required, the train control curve is sent to the train control system to control the operation of the freight train.

[0038] Compared with the prior art, the technical solution of the present invention has the following advantages and beneficial effects:

[0039] This solution uses precise operation status prediction technology to provide accurate information for the integration of freight railway scheduling and control, which helps to achieve precise adjustment of train operation plans and refinement of train control.

[0040] The train operation planning module of this solution uses an LSTM neural network system to accurately simulate the train's operating status, which is more in line with the modeling requirements of the nonlinear system of freight trains. It can thus accurately predict the train's operating trajectory and remaining operating time, and then form the train's operation plan and route strategy.

[0041] This solution uses an expert system's freight train operation planning algorithm and combines the characteristics of freight trains to carefully plan the three stages of freight train operation and different scenarios, thereby predicting the train's future operating status information. Ultimately, it can accurately determine the train's remaining operating time, form an operation plan, and feed it back to the dispatching system to ensure train operation efficiency and smoothness. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 This is a schematic diagram of a freight train dispatching and control system applied to a train operation control system according to an embodiment of the present invention;

[0043] Figure 2 This is a logic diagram of a train operation status prediction system according to an embodiment of the present invention;

[0044] Figure 3 This is a flow chart of a method for dispatching and controlling a freight train according to an embodiment of the present invention;

[0045] Figure 4 This is a schematic diagram of the three-stage division of freight train control by the train operation planning module of an embodiment of the present invention. DETAILED DESCRIPTION

[0046] The following will be combined with the embodiment of the present invention Figures 1 to 4 , the technical solutions, structural features, objectives achieved and effects in the embodiments of the present invention are described in detail.

[0047] It should be noted that the drawings are in a very simplified form and use non-precise proportions. They are only used to conveniently and clearly assist in explaining the embodiments of the present invention, and are not used to limit the conditions for the implementation of the present invention. Therefore, they have no substantive technical significance. Any structural modification, change in proportional relationship or adjustment of size should still fall within the scope of the technical content disclosed in the present invention without affecting the efficacy and purpose that can be achieved by the present invention.

[0048] It should be noted that, in the present invention, relational terms such as first and second, etc. are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus comprising a series of elements includes not only the elements explicitly listed, but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus.

[0049] The present invention proposes a dispatching and control system and method for freight trains. By combining the current operating status of the freight train and the current route conditions, the operating situation within a certain distance ahead of the freight train is predicted. The system and method of the present invention can still provide relatively accurate guidance for the operation control of freight trains when faced with different performances of different freight trains and changes in weather conditions, thereby improving the efficiency and safety of train operation control.

[0050] The technical solution of the present invention will be described in detail below with reference to the accompanying drawings and embodiments.

[0051] See also Figure 1, shows a schematic diagram of a freight train dispatching and control system according to this embodiment applied to a train operation control system. The freight train dispatching and control system includes an information interface device 11, a train operation status prediction system 12, and a dispatching system 13. The information interface device 11 is connected to the trackside dispatching equipment of the freight train track to obtain basic line data and environmental information of the line on which the train is located in real time. The information interface device 11 is also connected to the train operation status prediction system 12, and sends the basic line data and environmental information to the train operation status prediction system 12 for use in subsequent modeling and operation status prediction by the train operation status prediction system 12. The dispatching system 13 is connected to the train control data trackside interface device 22 and communicates bidirectionally. The train control data trackside interface device 22 is also connected to the train control system (train operation control system) 23 and communicates bidirectionally for data exchange. The train control system 23 is used to obtain the status of the freight train in real time and send train operation data to the train control data trackside interface device 22, so that the train control data trackside interface device 22 sends the train operation data to the dispatching system 13 in real time. The dispatching system 13 also communicates bidirectionally with the train operation status prediction system 12, transmitting train operation data to the system in real time for subsequent modeling and operation status prediction. When the train operation status prediction system 12 predicts train operation times and train control curves, it also transmits these to the dispatching system 13. After evaluation and assessment, the dispatching system 13 transmits these data to the train control system 23 via the train control data trackside interface device 22 to control freight train operations.

[0052] In this embodiment, the dispatching system 13 is a CTC (Centralized Traffic Control) system. The train control system 23 communicates with the train to obtain the status of the freight train in real time and controls the operation of the freight train after receiving the train control curve.

[0053] The train operation data includes train control data and train operation status data. The train control data includes information such as the train control system's control level, current, and speed. The train operation status data includes information such as train speed, acceleration, train position, and train formation. The basic line data includes line layout, slope, kilometer mark, speed limit information, etc. The environmental information includes fog, rain, and snow. Because environmental information may change with train operation, considering the impact of environmental factors on train operation control is very important for accurate situation prediction.

[0054] In this embodiment, the freight train operates in CO mode (CTCS (China Train Control System)-Level 0) and adopts the LKJ (Train Operation Monitoring Device) vehicle control system as the train control system 23. Therefore, it is necessary to set up a train control data trackside interface device 22, and set up an ATO (Automatic Train Operation System) automatic driving system (such as Shuohuang's LKJ-based automatic driving) on ​​the on-board end to communicate between the LKJ vehicle control system 23 and the dispatching system 13, so that the dispatching system 13 can obtain train operation data. In other embodiments, when in the moving block train control mode, ATP (Automatic Train Protection System) and ATO are used as the train control system 23, since there are trackside equipment between the dispatching system 13 and ATP and ATO, there is no need to set up additional information interaction equipment, and it is only necessary to increase the information interaction function between the dispatching system 13 and the train control system 23.

[0055] After acquiring the train operation data, the train operation status prediction system 12 stores the historical data of the train operation data for basic modeling, and uses the real-time data of the train operation data to dynamically adjust the modeling, and then predicts the subsequent operation status based on the adjusted model and the real-time train operation data.

[0056] Specifically, see Figure 2 The train operation status prediction system 12 includes a train characteristic identification module 121 and a train operation planning module 122. The train characteristic identification module 121 establishes a train operation control model based on the LSTM (long short-term memory network) neural network algorithm based on the basic line data, environmental information and the train operation data. The train operation planning module 122 predicts the train operation status based on the optimization strategy according to the train operation control model, the basic line data, environmental information and the train operation data, generates a train operation time and a train control curve, and sends them to the dispatching system 13. The train control curve is the target speed curve of the train and is the train operation strategy.

[0057] The dispatching system 13 determines whether the train running time conflicts with other train plans. If not, the dispatching system 13 sends the train control curve to the train control system 23 to control the train running.

[0058] The dispatching system 13 also communicates with the train control center (TCC) 21, and sends the train operation order, time and route strategy to it based on the train control curve, so that the TCC sends corresponding route control signals and track codes to the track to ensure that subsequent freight trains can pass or stop normally.

[0059] The freight train dispatching and control method of this embodiment adopts the above-mentioned freight train dispatching and control system, such as Figure 3As shown, the specific steps include:

[0060] S1. The information interface device 11 obtains the basic line data and environmental information of the train line in real time and sends it to the train operation status prediction system 12; the dispatching system 13 exchanges data with the train control system 23 to obtain the train operation data in real time and sends it to the train operation status prediction system 12;

[0061] S2, the train characteristic identification module 121 in the train operation status prediction system 12 establishes a train operation control model based on the LSTM neural network algorithm according to the basic line data, environmental information and the train operation data;

[0062] S3. The train operation planning module 122 in the train operation status prediction system 12 performs prediction calculations on the train operation status based on the train operation control model, the basic line data, the environmental information and the train operation data based on an optimization strategy, generates a train operation time and a train control curve, and sends the result to the dispatching system 13.

[0063] S4. The dispatching system 13 determines whether the train running time is reasonable. If it is reasonable, the operation plan will be adjusted according to the train running time, and the train control curve will be sent to the train control system 23 to control the operation of the freight train.

[0064] Wherein, the step S2 further includes the following steps:

[0065] S21, based on the basic line data, environmental information and historical data of the train operation data, the train characteristic identification module 121 establishes a basic train operation control model based on the LSTM neural network algorithm;

[0066] S22. Apply the train operation control basic model to the train operation status prediction system 12, and dynamically adjust the train operation control basic model according to the real-time data of the basic line data, environmental information and train operation data to obtain the final train operation control model.

[0067] In the above steps, considering that freight trains have many types of locomotives, complex traction and braking systems, and uncertain formations compared to other trains, the use of traditional transfer function methods will result in difficult, complex, and inaccurate modeling. Therefore, this embodiment establishes a basic model of train operation control based on a neural network. The neural network model has the advantages of powerful nonlinear modeling capabilities, self-learning and self-adaptation capabilities, and is more in line with the modeling requirements of nonlinear systems such as freight trains.

[0068] The train operation control model includes parameters of the train traction and braking systems. By establishing such a train operation control model, the actual performance of the train in operation can be simulated, so that the train operation status can be accurately predicted subsequently.

[0069] In step S21, the train characteristic identification module 121 uses the slope and kilometer mark in the basic line data, environmental information, the control level, current, speed and other information in the train control data, and the speed and acceleration information in the train operation status data to offline generate a basic train operation control model based on the LSTM neural network algorithm.

[0070] In step S22, the train operation control basic model is applied to the train operation status prediction system 12, and basic line data, environmental information, train control data and train operation status data are received in real time. The train operation control basic model is dynamically adjusted. During the process, the train operation control basic model is used to predict the train speed and acceleration, and then the prediction result of the train operation control basic model is compared with the actual train operation data. If the difference between the two is within the threshold range, the train operation control basic model at this time is used as the train operation control model.

[0071] The LSTM neural network algorithm used in this embodiment includes a forget gate, an input gate, a cell state update, and an output gate. The formula and meaning are as follows:

[0072] f t =σ(W hf x t +W hf h t-1 +b f )

[0073] i t =σ(W hi x t +W hi h t-1 +b i )

[0074]

[0075] o t =σ(W ho x t +W ho h t-1 +b o )

[0076] h t =o t *tanh(C t )

[0077] Among them, ft It represents the output of the forget gate at time t, which is a vector between 0 and 1, used to decide which information to discard from the cell state, that is, the factors that are not related to the train traction and braking parameters. hf is the weight matrix corresponding to the forget gate, which is the parameter setting for linear transformation of input information, b f is the bias vector corresponding to the forget gate, which is used to adjust the result after linear transformation to better adapt to the characteristics of train driving data. t is the current input, h t-1 is the hidden state at time t-1, σ is the Sigmoid activation function, which maps the result after linear transformation to the range of 0 to 1 to ensure that the output forgetting degree value meets the requirements; i t is the output of the input gate at time t, which is also a vector between 0 and 1, and determines how many new train driving factors can be updated to the cell state. hi and b i are the weight matrix and bias vector corresponding to the input gate respectively; is the candidate cell state; C t is the updated cell state at time t, f t *C t-1 Is to use the output f of the forget gate t For the cell state C at the previous moment t-1 Perform selective forgetting, that is, multiplying elements by element to remove the information that needs to be discarded in the cell state at the previous moment (that is, information that is irrelevant to the train's running status). The output i of the input gate is used t and candidate cell states Perform element-wise multiplication, update the new information to the cell state according to the ratio determined by the input gate, and finally add the two to complete the update of the cell state from time t-1 to time t, so that the cell state can retain important long-term memory and integrate new information at the current moment; t is the output of the output gate at time t, which is also a vector between 0 and 1. It determines which information in the cell state can be output to the hidden state at the current moment. W ho and b o are the weight matrix and bias vector corresponding to the output gate respectively.

[0078] Based on the aforementioned LSTM neural network algorithm formula, a basic train operation control model is established. Upon receiving real-time train operation data, it is iteratively adjusted until the difference between the train speed and acceleration predicted by the established train operation control model and the actual train operation data falls within a threshold range. The threshold value can be determined based on the accuracy requirements. Because the LSTM neural network algorithm is widely used and mature, the algorithm's iterative process is not detailed here.

[0079] Furthermore, after step S22, the prediction results of the train operation control model and the actual train operation data are continuously monitored. If the difference between the two exceeds the threshold range, the train operation control model is continuously dynamically adjusted until the difference between the two meets the threshold requirement again.

[0080] After the above adjustments, the final train operation control model can correct the performance differences between different freight trains due to changes in composition, wear and tear, and load. It can also reflect actual environmental changes, such as the impact of wind, rain and snow on train control.

[0081] In step S3, the train operation planning module 122 divides the travel process of the freight train into the station start-up and departure stage, the station parking and braking stage and the interval operation stage. The interval operation stage also includes a variety of scenarios including long downhill slopes, long uphill slopes, over-phase, deceleration areas and constant speed areas. Then, according to different scenarios and stages, based on the train operation control model, basic line data, environmental information and real-time train operation data, the optimal travel strategy within a certain distance ahead of the train is comprehensively calculated.

[0082] The formula for the optimization strategy is

[0083]

[0084] Among them, J is the objective function, w1 is the weight of train traction energy consumption, F k is the traction force of the train per step, N is the total step length in the train section, Δs is the length of each step, w2 is the weight of the train following the target speed, which is directly related to the train section running time, E k is the kinetic energy of the train per step, m is the mass of the train in kg, v desk is the following speed of the train per step, in m / s. For energy saving considerations, For efficiency (time) considerations, the optimization strategy of this embodiment takes both train operation efficiency and energy saving into consideration.

[0085] During the calculation, the optimal objective function J is calculated under the equality constraints of the dynamic equation and the upper and lower limit constraints of the traction force F, the braking force B, and the kinetic energy E; the equality constraints of the dynamic equation are E k =E k-1 +F k Δs-B k ·Δs-(W i,k +W r,k +W a,k )·Δs, where W i,k is the slope resistance, B k is the braking force, W r,k is the curve resistance, W a,k is the basic resistance of train operation; the upper and lower limits of the traction force F, braking force B, and kinetic energy E are: 0≤F≤F max , 0≤B≤B(E), 0≤E≤E desk , where F max is the maximum traction force that the train can provide during operation, B(E) is the braking force that the train can provide under the current kinetic energy, and B is expressed as a linear function of E, that is, B = αE + β, E desk It is the maximum kinetic energy of the train during operation.

[0086] After calculating the optimized objective function J, the ideal data of the train operation such as traction and speed corresponding to the optimized objective function J can be obtained. Then, the train operation control model of the train characteristic identification module 121 is used to calculate the future operation status information of the train based on the ideal train operation data. Finally, the remaining operation time of the train to arrive at the next station, the corresponding operation plan, route strategy and operation trajectory can be calculated.

[0087] However, if Figure 4As shown in the figure, each scenario and stage has different focuses. The station start-up and departure stage is divided into two stages: low-speed start with fixed traction operation and dynamic planning acceleration. The station parking brake stage is also divided into two stages: deceleration process of one braking and slow control of low-speed parking. The five scenarios in the interval operation stage have different focuses. Long downhill slopes require determining the position where the train starts to apply air brakes, while considering the entrance speed. When the train passes the long downhill slope in a cyclic braking control mode, the braking system in the train operation control model needs to be used to predict the position and speed of the train leaving the downhill slope. Long uphill slopes focus on the train's dynamic performance and climbing ability. The entrance speed before the uphill slope needs to be calculated to ensure that the train carries enough kinetic energy to overcome the influence of gravity. The traction system in the train operation control model is used to predict the exit speed of the train leaving the long uphill slope area. The over-phase area focuses on the train's minimum entry speed and exit speed. The deceleration area needs to determine the starting position speed of one braking and the speed position after deceleration. The constant speed area connects the previous four scenarios together through planning.

[0088] Because each of the aforementioned scenarios and stages has different focus points, when predicting the train's operating status, the train operation planning module 122 first needs to consider the scenarios and stages ahead of the train. It then calculates the objective function J for each stage and scenario within a certain distance ahead of the train, and then sums these as the final objective function J'. The module then calculates the parameters corresponding to the optimized final objective function J'. Finally, based on the parameters of the final objective function J', the corresponding train operating time and train control curve are calculated. The calculated parameters of the final objective function J' generally include speed, acceleration, deceleration position, acceleration position, etc. for each stage and scenario. Based on these parameters, the train operating time required for the train to reach the next station and the corresponding train control curve can be calculated.

[0089] The specific distance of the train ahead depends on the actual situation. For example, if the train needs to stop at the next station, the ahead distance is the distance between the train and the next station. If the train does not stop at the next station but only passes through, the ahead distance can be the distance between the train and the next station, the distance between the train and the next station, or the distance between the train and any location between the next station and the next station. Although the ahead distance can also be the distance between the train and a location at the next station, since trains generally can only stop within stations, it is more practical to predict the time when the train arrives at or passes through the station.

[0090] In step S4, when the train operation planning module 122 generates the final train operation time and train control curve and sends them to the dispatching system 13, if the dispatching system 13 determines that the train operation time conflicts with other train plans and is unreasonable, that is, the arrival time of this freight train at the station coincides with that of other trains, then delay information is sent to the train operation planning module 122; the delay information can be only information to notify the train operation planning module 122 to extend the time, or it can be a specific extension time, such as 30 seconds, 50 seconds, etc., to avoid conflicts with other train operation plans.

[0091] When the train operation planning module 122 receives the delay information, it formulates a new train operation time and train control curve based on the train operation control model, the basic line data, environmental information, the train operation data and the delay information, and sends them to the dispatching system 13. The dispatching system 13 again determines whether the new train operation time and train control curve are reasonable; repeat this process until the dispatching system 13 determines that the train operation time is reasonable.

[0092] Among them, when the train operation planning module 122 formulates a new train operation time and train control curve, if there is a constant speed zone scene within a certain distance ahead of the train, the parameters under the constant speed zone scene are adjusted first, that is, the constant speed, and then the parameters in other stages and scenes are adjusted based on the constant speed to improve efficiency and operation smoothness.

[0093] Finally, the dispatching system 13 adjusts the operation plan according to the reasonable train running time, and sends the train control curve corresponding to the final train running time to the train control system 23. The train control system 23 controls the train operation based on the train control curve.

[0094] This invention integrates the synergistic coupling and deep functional integration of existing freight railway dispatching and train control systems, achieving information and collaborative integration of the dispatching and train control systems based on networking, intelligence, and digitization. This solution enables accurate prediction of freight trains, providing important support for precise adjustment of dispatch plans and fine-grained control of trains.

[0095] Although the present invention has been described in detail through the above preferred embodiments, it should be understood that the above description is not intended to limit the present invention. After reading the above description, various modifications and substitutions of the present invention will become apparent to those skilled in the art. Therefore, the scope of protection of the present invention should be defined by the appended claims.

Claims

1. A method for dispatching and controlling freight trains, characterized in that: The following steps are involved: S1. The information interface device obtains basic line data and environmental information of the train line in real time and sends it to the train operation status prediction system; the dispatching system and the train control system exchange data to obtain train operation data in real time and send it to the train operation status prediction system; S2. The train characteristic identification module in the train operation status prediction system establishes a train operation control model based on the LSTM neural network algorithm according to the basic line data, environmental information and the train operation data. The train operation control model includes parameters of the train traction and braking systems; S3. The train operation planning module in the train operation status prediction system predicts the train operation status based on the train operation control model, the basic line data, the environmental information and the train operation data based on the optimization strategy, generates the train operation time and train control curve, and sends it to the dispatching system; S4. The dispatching system determines whether the train running time is reasonable. If so, the dispatching system adjusts the running plan according to the train running time and sends the train control curve to the train control system to control the operation of the freight train. In step S4, when the dispatching system determines that the train running time conflicts with other train plans, it sends delay information to the train operation planning module; The train operation planning module formulates a new train operation time and train control curve based on the train operation control model, the basic line data, the environmental information, the train operation data and the delay information, and sends the new train operation time and train control curve to the dispatching system; repeating this process until the dispatching system determines that the train operation time is reasonable; In step S3, the formula of the optimization strategy is: Among them, J is the objective function, w1 is the weight of train traction energy consumption, F k is the traction force of the train per step, N is the total step length in the train section, Δs is the length of each step, w2 is the weight of the train following the target speed, which is directly related to the train section running time, E k is the kinetic energy of the train per step, m is the mass of the train in kg, v desk is the following speed of the train per step, in m / s; The train operation planning module divides the freight train's travel process into a station start-up phase, a station parking brake phase, and a section operation phase. The section operation phase also includes various scenarios such as long downhill slopes, long uphill slopes, over-phase, deceleration zones, and constant speed zones. In step S3, the train operation planning module calculates the objective function J in each stage and scenario within a certain distance ahead of the train, and then takes the sum of the objective functions as the final objective function J', calculates the parameters corresponding to the optimized final objective function J', and finally calculates the corresponding train operation time and train control curve; When the train operation planning module formulates a new train operation time and train control curve, if there is a constant speed zone scenario within a certain distance ahead of the train, the parameters in the constant speed zone scenario are adjusted first, and then the parameters in other stages and scenarios are adjusted.

2. The method for dispatching and controlling a freight train according to claim 1, wherein: The step S2 specifically includes: S21, based on the basic line data, environmental information and historical data of the train operation data, the train characteristic identification module establishes a basic train operation control model based on the LSTM neural network algorithm; S22. Apply the train operation control basic model to the train operation status prediction system, and dynamically adjust the train operation control basic model according to the basic line data, environmental information and real-time data of the train operation data to obtain the train operation control model.

3. The method for dispatching and controlling a freight train according to claim 2, wherein: In step S22, during the process of dynamically adjusting the train operation control basic model, the prediction result of the train operation control basic model is compared with the actual train operation data. If the difference between the two is within the threshold range, the train operation control basic model at this time is used as the train operation control model.

4. The method for dispatching and controlling a freight train according to claim 3, wherein: After step S22, the prediction result of the train operation control model and the actual train operation data are continuously monitored. If the difference between the two exceeds the threshold range, the train operation control model is continuously dynamically adjusted.

5. The method for dispatching and controlling a freight train according to claim 1, wherein: In step S3, the train operation planning module calculates the optimized objective function J based on the train operation control model and the real-time data of the train operation data and the formula of the optimization strategy, and then obtains its corresponding parameters to calculate the corresponding train operation time and train control curve.

6. The method for dispatching and controlling a freight train according to claim 5, wherein: In step S3, the optimized objective function J is calculated under the equality constraints of the dynamic equation and the upper and lower limit constraints of the traction force F, the braking force B, and the kinetic energy E; The equality constraint of the kinetic equation is E k =E k-1 +F k Δs-B k ·Δs-(W i,k +W r,k +W a,k )·Δs, where W i,k is the slope resistance, B k is the braking force, W r,k is the curve resistance, W a,k It is the basic resistance of train operation; The upper and lower limits of the traction force F, braking force B, and kinetic energy E are: 0≤F≤F max , 0≤B≤B(E), 0≤E≤E desk , where F max is the maximum traction force that the train can provide during operation, B(E) is the braking force that the train can provide under the current kinetic energy, and E desk It is the maximum kinetic energy of the train during operation.

7. The method for dispatching and controlling a freight train according to claim 1, wherein: The train operation data includes train control data and train operation status data. The train control data includes the operation level, current, and speed of the train control system, and the train operation status data includes the train speed and acceleration.

8. A freight train dispatching and control system, used to implement the freight train dispatching and control method according to any one of claims 1 to 7, characterized in that: include: Information interface equipment, train operation situation prediction system and dispatching system; The information interface device is used to obtain basic line data and environmental information and send it to the train operation status prediction system; The train operation status prediction system is also connected to the dispatching system, and the dispatching system sends train operation data to the train operation status prediction system; The train operation situation prediction system includes a train characteristics identification module and a train operation planning module; The train characteristic identification module establishes a train operation control model based on the LSTM neural network algorithm according to the basic line data, environmental information and the train operation data; The train operation planning module predicts the train operation status based on the train operation control model, the basic line data, the environmental information and the train operation data based on the optimization strategy, generates the train operation time and train control curve, and sends it to the dispatching system; The dispatching system also interacts with the train control system data to obtain the train operation data in real time; the dispatching system is also used to determine whether the train operation time needs to be adjusted. If no adjustment is required, the train control curve is sent to the train control system to control the operation of the freight train.

Images