A speed tracking control method for heavy-load trains used in long downhill cyclic braking

Abstract

The present invention discloses a heavy-load train speed tracking control method for long downhill cyclic braking. The heavy-load train speed tracking controller is designed based on a disturbance observer and an adaptive iterative learning method. The information during the train operation is effectively used to perform controller parameter adaptation. The timing of applying air braking does not rely on the accurate modeling of the heavy-load train air brake system and does not rely on empirical timing, thereby achieving accurate tracking control of the train speed and position. The present invention provides a control decision judgment standard for cyclic braking and relief of heavy-load trains when running on long downhill slopes, thereby ensuring the safety of train operation.

Description

Technical Field

[0001] The invention belongs to the technical field of train operation, and in particular relates to a design of a speed tracking control method for a heavy-load train used for long downhill cyclic braking. Background Art

[0002] The characteristics of heavy-duty trains with extra-long bodies and extra-large loads pose great challenges to drivers' safe driving. In particular, China's existing 20,000-ton heavy-duty lines all contain long downhill slopes, which require a cyclic braking strategy that combines electric braking with air braking. The timing of each braking and relief is heavily dependent on the driver's operating experience, and the operating differences are significant, which affects the safe and efficient operation of the train. Therefore, considering the actual operating environment of heavy-duty trains, designing an adaptive speed tracking controller to realize online automatic control of trains is of great significance to the development of heavy-duty railways.

[0003] The operation control technology of heavy-load trains has attracted the attention of many domestic scholars. Considering the current development status of heavy-load trains in my country, a large number of studies have considered the application and relief of air brakes as discrete variables, and used the force analysis or speed limit of the train on a long downhill slope as the judgment condition for the timing of cyclic braking application to achieve train optimization control. Such methods require accurate modeling of heavy-load trains, but since the train operation process is often affected by a variety of time-varying external environments, it is impossible to obtain an accurate model, and complex models often reduce the solution efficiency, making it difficult to achieve real-time online control of the train.

[0004] The operation characteristics of heavy-load trains in long downhill cyclic braking are highly repeatable, which is manifested in repeated lines, repeated dynamic model structures, and similar operation strategies. Adaptive Iterative Learning Control (AILC) can combine the advantages of adaptive control and iterative learning control, make full use of the repeatability of train operation, effectively deal with system parameters and model uncertainties, and achieve fast, high-precision, and complete tracking of the desired trajectory. At present, AILC control research in the field of rail transit is mainly aimed at high-speed trains, considering unknown time-varying resistance and measurement noise in train operation. AILC has not yet involved heavy-load trains. Since heavy-load trains are subject to high / low speed limits, traction / braking characteristics, and other constraints during operation, the cyclic braking process is also subject to relief time constraints. Many constraints need to be handled when designing the controller. In addition, the control force of heavy-load trains has long lag and high response delay. Simple adaptive iterative learning control cannot update parameters in time. There are huge longitudinal impulses in the operation of heavy-load trains, so it is difficult to achieve good tracking effects. The disturbance observer can timely feedback the difference between the actual model and the nominal model and timely reflect the state of the controlled object. Adding a disturbance observer to the speed tracking controller of a heavy-load train can reduce the impact of the high response delay of the heavy-load train and improve the timeliness of the adaptive parameter update. Existing research mainly combines the disturbance observer with traditional sliding mode control and uses the error obtained by the disturbance observer to compensate the control. Summary of the invention

[0005] The purpose of the present invention is to propose a speed tracking control method for a heavy-load train used for long downhill cyclic braking, which realizes accurate tracking control of the train speed and position under the conditions of unknown disturbances, unknown line conditions, limited train status and limited input during the operation of the heavy-load train.

[0006] The technical solution of the present invention is: a speed tracking control method for a heavy-load train for long downhill cyclic braking, comprising the following steps:

[0007] S1. Obtain the basic data, initial control data and target running speed curve of the heavy-load train.

[0008] S2. Calculate the tracking error based on the basic data of the heavy-load train, the control data of the previous operation process and the target operating speed curve of the train.

[0009] S3. Adjust the controller parameters of the heavy-load train according to the tracking error.

[0010] S4. Apply the adjusted controller to the heavy-load train operation control system and record the control data for the next controller parameter adjustment.

[0011] Furthermore, the basic data of the heavy-load train in step S1 include the train curb weight M, basic running resistance F b (v) Train traction force F t (v) and train electric braking force F d (v), the initial control data includes the initial speed v0(t), the initial position s0(t) and the initial control amount u0(t), where v represents speed and t represents time.

[0012] Furthermore, the calculation formula of the tracking error in step S2 is:

[0013] e k =e vk +λe sk

[0014] e vk =v r (t)-v k-1 (t)

[0015] e sk =s r (t)-s k-1 (t)

[0016] where e k represents the composite tracking error during the kth operation, e vk represents the speed tracking error during the kth operation, e sk represents the position tracking error during the kth operation, λ represents the weight coefficient of speed and position tracking performance, and v r (t) represents the reference speed provided by the train target running speed curve, v k-1 (t) represents the train speed during the k-1th operation, s r (t) represents the reference position provided by the train target running speed curve, s k-1 (t) represents the train position during the k-1th operation, k = 1, 2, 3, ...

[0017] Furthermore, the control data of the heavy-load train during the first operation is obtained by solving the initial traction calculation model of the train. The initial traction calculation model of the train is specifically:

[0018]

[0019] Where x represents the position, represents the derivative of x, represents the derivative of v, F m Indicates the train air braking force, F a (x) represents the additional resistance of the train line.

[0020] Further, step S3 includes the following sub-steps:

[0021] S31. Design a train traction calculation model.

[0022] S32. Adjust the adaptive iterative learning parameters of the heavy-load train controller according to the tracking error and the train traction calculation model.

[0023] S33, judging whether air brake intervention is required in the heavy-load train controller.

[0024] S34. Adjust the disturbance observer parameters of the heavy-load train controller.

[0025] S35. Adjust the compound controller parameters.

[0026] Furthermore, the train traction calculation model in step S31 is specifically:

[0027]

[0028] where u k (t) represents the train control quantity during the k-th operation, represents the unknown parameter vector during the kth run, a k (t),b k (t),c k (t) are the basic resistance parameters of train operation, F ak (t) represents the additional resistance of the train running line during the kth operation process, represents the observable data vector during the k-th run, σ k (t) represents the unknown disturbance during the kth run.

[0029] Further, step S32 includes the following sub-steps:

[0030] S321. Set the speed limit constraint for heavy-load trains, the starting and ending kilometer mark constraints for heavy-load trains, and the input constraints for the heavy-load train control system:

[0031] v min,k (t)≤v k (t)≤v max,k (t)

[0032] s min,k (t)≤s k (t)≤s max,k (t)

[0033] u' min (t)≤u' k (t)≤u' max (t)

[0034] where u'k (t) represents the traction / electric braking force control amount during the kth operation, u' min (t),u' max (t) represent the upper and lower limits of the traction / electric braking force input of the heavy-duty train, v min,k (t),v max,k (t) represent the lower speed limit and upper speed limit of heavy-load train respectively, s min,k (t),s max,k (t) represent the starting and ending kilometer marks of heavy-load train respectively.

[0035] S322, according to the heavy-load train operation data obtained by solving the train traction calculation model, the adaptive iterative learning parameters are obtained by satisfying the constraint conditions of step S321:

[0036]

[0037]

[0038] where u k (t) = u' k (t)-u” k (t)·F m ,u" k (t) represents the air brake force control coefficient during the k-th operation, α>0 represents the feedback control gain constant, β>0 represents the parameter adaptive update constant, represents the adaptive intermediate parameter, represents the barrier Lyapunov function, k a >0 is the designed boundary value, represents the parameter vector saturation projection operator, and has:

[0039]

[0040] in represents the maximum value of the parameter vector change, Represents the minimum value of the change in the parameter vector.

[0041] Further, step S33 includes the following sub-steps:

[0042] S331, introduce air brake relief timer to record the relief time T air , define the initial state T air =200.

[0043] S332, when the train is air-braked, that is, u″ k When (t) = 1, record the relief time T air =0.

[0044] S333, when the train operating condition changes from air braking to relief, that is, u″ k (t)=1→u″ k When (t) = 0, the air brake release counter starts timing and records the release time T air .

[0045] S334, when T air >180,u' k (t-1)+Δu' k (t-1)<-F d,max And v max -5 <v k (t) <v max When u″ k (t) = 1, air brake intervenes to supplement braking force, otherwise u″ k (t) = 0, relief is performed; where u' k (t-1) represents the traction / electric braking force control amount of the previous step, Δu' k (t-1) represents the change in the traction / electric braking force control amount in the next step obtained based on the calculation result of the previous step, F d,max Indicates the maximum value of the electric braking force, v max Indicates the maximum speed of a heavy-load train.

[0046] Furthermore, in step S34, a sliding mode disturbance observer is added to the heavy-load train controller to adjust disturbance observer parameters of the heavy-load train controller. The sliding mode disturbance observer is specifically:

[0047]

[0048] in represents the train speed observation value obtained by the observer, express The derivative of represents the train disturbance observation value obtained by the observer, w σ represents the reference gain value of the sliding mode control function, represents the velocity observation error, represents the sliding mode control function, and:

[0049]

[0050] Where μ represents a constant to weaken the chattering phenomenon of the sliding mode disturbance observer, represents the perturbation observation error, k v represents the switching gain factor, k εv represents the velocity coefficient to reach the sliding surface, represents the sliding surface function.

[0051] Furthermore, in step S35, the composite controller is specifically:

[0052]

[0053] Among them U k represents a composite controller, It represents the saturation projection operator of the total control quantity of the train, which satisfies the constraints of the actual traction output and braking force output of the train:

[0054]

[0055] where F t,max (v) represents the maximum traction force of the train, F d,max (v) represents the maximum electric braking force of the train.

[0056] The beneficial effects of the present invention are:

[0057] (1) The present invention effectively utilizes information from the train operation process to perform controller parameter adaptation. The timing of applying the air brake does not rely on the precise modeling of the heavy-load train air brake system or empirical timing, thereby achieving precise tracking and control of the train speed and position.

[0058] (2) The present invention provides a control decision-making criterion for the cyclic braking and relief of heavy-load trains when running on long downhill slopes, thereby ensuring the safety of train operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 The figure shows a flow chart of a speed tracking control method for a heavy-load train for long downhill cyclic braking provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0060] Now, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings. It should be understood that the embodiments shown and described in the accompanying drawings are only exemplary and are intended to explain the principles and spirit of the present invention, rather than to limit the scope of the present invention.

[0061] The embodiment of the present invention provides a speed tracking control method for a heavy-load train for long downhill cyclic braking, such as Figure 1 As shown, the following steps S1 to S4 are included:

[0062] S1. Obtain the basic data, initial control data and target running speed curve of the heavy-load train.

[0063] In the embodiment of the present invention, the basic data of the heavy-load train include the train curb weight M, the basic running resistance F b (v) Train traction force F t (v) and train electric braking force F d(v). The initial control data includes the initial speed v0(t), the initial position s0(t) and the initial control amount u0(t), where v represents the speed and t represents the time. The target running speed curve of the train is obtained from the operation diagram.

[0064] S2. Calculate the tracking error based on the basic data of the heavy-load train, the control data of the previous operation process and the target operating speed curve of the train.

[0065] In the embodiment of the present invention, the tracking error is the difference between the actual running speed and position of the heavy-load train and the target running speed curve, and the calculation formula is:

[0066] e k =e vk +λe sk

[0067] e vk =v r (t)-v k-1 (t)

[0068] e sk =s r (t)-s k-1 (t)

[0069] where e k represents the composite tracking error during the kth operation, e vk represents the speed tracking error during the kth operation, e sk represents the position tracking error during the kth operation, λ represents the weight coefficient of speed and position tracking performance, which is a positive constant, and v r (t) represents the reference speed provided by the train target running speed curve, v k-1 (t) represents the train speed during the k-1th operation, s r (t) represents the reference position provided by the train target running speed curve, s k-1 (t) represents the train position during the k-1th operation, k = 1, 2, 3, ...

[0070] In the embodiment of the present invention, the control data of the last operation process includes the train speed v of the last operation process. k-1 (t), train position s k-1 (t) and train control quantity u k-1 (t). The control data of the heavy-load train during its first operation is obtained by solving the initial traction calculation model of the train. The initial traction calculation model of the train does not consider disturbance, that is, σ=0. The specific formula is obtained based on the basic data of the heavy-load train:

[0071]

[0072] Where x represents the position, represents the derivative of x, represents the derivative of v, F m Indicates the train air braking force, F a (x) represents the additional resistance of the train line.

[0073] S3. Adjust the controller parameters of the heavy-load train according to the tracking error.

[0074] Step S3 includes the following sub-steps S31 to S35:

[0075] S31. Design train traction calculation model:

[0076]

[0077] where u k (t) represents the train control quantity during the k-th operation, represents the unknown parameter vector during the kth run, a k (t),b k (t),c k (t) are the basic resistance parameters of train operation, F ak (t) represents the additional resistance of the train running line during the kth operation process, represents the observable data vector during the k-th run, σ k (t) represents the unknown disturbance during the kth run.

[0078] S32. Adjust the adaptive iterative learning parameters of the heavy-load train controller according to the tracking error and the train traction calculation model.

[0079] Step S32 includes the following sub-steps S321-S322:

[0080] S321. Set the speed limit constraint for heavy-load trains, the starting and ending kilometer mark constraints for heavy-load trains, and the input constraints for the heavy-load train control system:

[0081] v min,k (t)≤v k (t)≤v max,k (t)

[0082] s min,k (t)≤s k (t)≤s max,k (t)

[0083] u' min (t)≤u' k (t)≤u' max (t)

[0084] where u'k (t) represents the traction / electric braking force control amount during the kth operation, u' min (t),u' max (t) represent the upper and lower limits of the traction / electric braking force input of the heavy-duty train, v min,k (t),v max,k (t) represent the lower speed limit and upper speed limit of heavy-load train respectively, s min,k (t),s max,k (t) represent the starting and ending kilometer marks of heavy-load train respectively.

[0085] S322, according to the heavy-load train operation data obtained by solving the train traction calculation model, the adaptive iterative learning parameters are obtained by satisfying the constraint conditions of step S321:

[0086]

[0087]

[0088] where u k (t) = u' k (t)-u” k (t)·F m ,u" k (t) represents the air brake force control coefficient during the k-th operation, α>0 represents the feedback control gain constant, β>0 represents the parameter adaptive update constant, represents the adaptive intermediate parameter, represents the barrier Lyapunov function, k a >0 is the design boundary value, when the error e k tends to k a hour, tends to infinity, so e k ∈(-k a ,k a ) is a bounded value. represents the parameter vector saturation projection operator, and has:

[0089]

[0090] in represents the maximum value of the parameter vector change, Indicates the minimum value of the parameter vector change. In the embodiment of the present invention, the basic resistance coefficient a k (t), b k (t), c k The upper and lower limits of (t) are empirical values ​​obtained according to the "Train Traction Calculation Regulations", and the additional line resistance F akThe upper and lower limits of (t) are calculated based on the maximum slope, minimum curve radius and other parameters specified in the Railway Alignment Design.

[0091] S33, judging whether air brake intervention is required in the heavy-load train controller.

[0092] Step S33 includes the following sub-steps S331 to S334:

[0093] S331, introduce air brake relief timer to record the relief time T air , define the initial state T air =200.

[0094] S332, when the train is air-braked, that is, u″ k When (t) = 1, record the relief time T air =0.

[0095] S333, when the train operating condition changes from air braking to relief, that is, u″ k (t)=1→u″ k When (t) = 0, the air brake release counter starts timing and records the release time T air .

[0096] S334, when T air >180,u' k (t-1)+Δu' k (t-1)<-F d,max And v max -5 <v k (t) <v max When u″ k (t) = 1, air brake intervenes to supplement braking force, otherwise u″ k (t) = 0, relief is performed; where u' k (t-1) represents the traction / electric braking force control amount of the previous step, Δu' k (t-1) represents the change in the traction / electric braking force control amount in the next step obtained based on the calculation result of the previous step, F d,max Indicates the maximum value of the electric braking force, v max Indicates the maximum speed of a heavy-load train.

[0097] S34. Adjust the disturbance observer parameters of the heavy-load train controller.

[0098] In the embodiment of the present invention, in order to improve the anti-interference ability and tracking accuracy of the system, compensate for the influence of the long time lag of the heavy-load train response, and improve the timeliness of the adaptive parameter update, a sliding mode disturbance observer is added to the heavy-load train controller:

[0099]

[0100] in represents the train speed observation value obtained by the observer, express The derivative of represents the train disturbance observation value obtained by the observer, w σ represents the reference gain value of the sliding mode control function, represents the velocity observation error, represents the sliding mode control function, and:

[0101]

[0102] Where μ represents a constant to weaken the chattering phenomenon of the sliding mode disturbance observer, represents the perturbation observation error, k v represents the switching gain factor, k εv represents the velocity coefficient to reach the sliding surface, represents the sliding surface function.

[0103] S35. Adjust the compound controller parameters.

[0104] In the embodiment of the present invention, the composite controller is specifically:

[0105]

[0106] Among them U k represents a composite controller, It represents the saturation projection operator of the total control quantity of the train, which satisfies the constraints of the actual traction output and braking force output of the train:

[0107]

[0108] where F t,max (v) represents the maximum traction force of the train, F d,max (v) represents the maximum electric braking force of the train.

[0109] S4. Apply the adjusted controller to the heavy-load train operation control system and record the control data for the next controller parameter adjustment.

[0110] Those skilled in the art will appreciate that the embodiments described herein are intended to help readers understand the principles of the present invention, and should be understood that the protection scope of the present invention is not limited to such specific statements and embodiments. Those skilled in the art can make various other specific variations and combinations that do not deviate from the essence of the present invention based on the technical revelations disclosed by the present invention, and these variations and combinations are still within the protection scope of the present invention.

Claims

1. A speed tracking control method for heavy-load trains used for long downhill cyclic braking, characterized in that: The following steps are involved: S1. Obtaining basic data, initial control data and target running speed curve of heavy-load train; The basic data of the heavy-load train includes the train curb weight M , basic operating resistance , Train traction and train electric braking force , the initial control data includes the initial speed , initial position and initial control amount , v Indicates speed, t Indicates time; S2, calculating the tracking error according to the basic data of the heavy-load train, the control data of the last operation process and the target operating speed curve of the train; S3. Adjust the controller parameters of the heavy-load train according to the tracking error. The following steps are included: S31. Design a train traction calculation model; S32, adjusting the adaptive iterative learning parameters of the heavy-load train controller according to the tracking error and the train traction calculation model; S33, judging whether air brake intervention is required in the heavy-load train controller; S34, adjusting the disturbance observer parameters of the heavy-load train controller; S35, adjusting the compound controller parameters; The step S32 includes the following sub-steps: S321. Set the speed limit constraint for heavy-load trains, the starting and ending kilometer mark constraints for heavy-load trains, and the input constraints for the heavy-load train control system: in Indicates k The traction / electric braking force control amount during the operation, Respectively represent the upper and lower limits of the traction / electric braking force input of the heavy-duty train, They represent the low speed limit and high speed limit for heavy-load trains respectively. They represent the starting and ending kilometer marks of heavy-load trains respectively. Indicates k The train speed during the run, Indicates k The train position during the run, ; S322, according to the heavy-load train operation data obtained by solving the train traction calculation model, the adaptive iterative learning parameters are obtained by satisfying the constraint conditions of step S321: in Indicates k The train control quantity during the operation, , Indicates k The air brake force control coefficient during the operation, Indicates the train air braking force, Indicates k The unknown parameter vector during the run, These are the basic resistance parameters of train operation. Indicates k The additional resistance of the train running line during the operation, Indicates k The observable data vector during the run, represents the feedback control gain constant, represents the parameter adaptive update constant, represents the adaptive intermediate parameter, Indicates k The composite tracking error during the run, represents the barrier Lyapunov function, is the design boundary value, represents the parameter vector saturation projection operator, and has: in represents the maximum value of the parameter vector change, Represents the minimum value of the parameter vector change; S4, applying the adjusted controller to the heavy-load train operation control system, and recording the control data for the next controller parameter adjustment; In step S34, a sliding mode disturbance observer is added to the heavy-load train controller to adjust the disturbance observer parameters of the heavy-load train controller. The sliding mode disturbance observer is specifically: in represents the train speed observation value obtained by the observer, express The derivative of represents the train disturbance observation value obtained by the observer, represents the reference gain value of the sliding mode control function, represents the velocity observation error, represents the sliding mode control function, and: in represents the constant to weaken the chattering phenomenon of the sliding mode disturbance observer, represents the perturbation observation error, represents the switch gain coefficient, represents the velocity coefficient to reach the sliding surface, represents the sliding surface function; The composite controller in step S35 is specifically: in represents a composite controller, It represents the saturation projection operator of the total control quantity of the train, which satisfies the constraints of the actual traction output and braking force output of the train: in Indicates the maximum traction of the train. Indicates the maximum electric braking force of the train.

2. The heavy-load train speed tracking control method according to claim 1, characterized in that: The calculation formula of the tracking error in step S2 is: in Indicates k The speed tracking error during the first run is: Indicates k The position tracking error during the run is Represents the weight coefficient of speed and position tracking performance, Indicates the reference speed provided by the train target operating speed curve. Indicates k -1 train speed during the run, Indicates the reference position provided by the train target running speed curve. Indicates k -1 Train position during the run.

3. The heavy-load train speed tracking control method according to claim 2, characterized in that: The control data of the heavy-load train during the first operation is obtained by solving the train initial traction calculation model, and the train initial traction calculation model is specifically: in x Indicates location, express x The derivative of express v The derivative of Indicates the additional resistance of the train line.

4. The heavy-load train speed tracking control method according to claim 3 is characterized in that: The train traction calculation model in step S31 is specifically: in, Indicates k Unknown disturbances during the run.

5. The heavy-load train speed tracking control method according to claim 4, characterized in that: The step S33 comprises the following sub-steps: S331, introduce air brake relief timer to record relief time , define the initial state ; S332, when the train is air-braked, that is Record the relief time ; S333, when the train operating condition is changed from air braking to relief, that is When the air brake release counter starts timing, the release time is recorded. ; S334, when , and hour, , air brake intervenes to supplement braking force, otherwise , to alleviate; Indicates the traction / electric braking force control amount of the previous step, Indicates the change in the traction / electric braking force control amount for the next step obtained based on the calculation result of the previous step. Indicates the maximum value of the electric braking force. Indicates the maximum speed of a heavy-load train.

Images