CN106154840A - A kind of asynchronous braking device and method of heavy-load combined train based on Model Predictive Control - Google Patents
A kind of asynchronous braking device and method of heavy-load combined train based on Model Predictive Control Download PDFInfo
- Publication number
- CN106154840A CN106154840A CN201610853319.8A CN201610853319A CN106154840A CN 106154840 A CN106154840 A CN 106154840A CN 201610853319 A CN201610853319 A CN 201610853319A CN 106154840 A CN106154840 A CN 106154840A
- Authority
- CN
- China
- Prior art keywords
- control
- module
- asynchronous
- braking
- locomotive
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B13/00—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion
- G05B13/02—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric
- G05B13/04—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators
- G05B13/042—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators in which a parameter or coefficient is automatically adjusted to optimise the performance
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Computation (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Electric Propulsion And Braking For Vehicles (AREA)
Abstract
The invention discloses a kind of asynchronous braking device and method of heavy-load combined train based on Model Predictive Control.Being different from traditional synchronous braking all locomotives in braking procedure to take consistent brake operating, asynchronous braking to consider pull strength etc. before and after train work information, rail conditions, locomotive to be braked operation, each locomotive is independently braked.Train is divided into several fence by landform according to car body fence thought by the present invention, obtains optimum reference locus by being predicted the reference locus of multiple fence;Utilize T S forecast model based on satisfaction, reference locus is carried out feedback compensation, obtain meeting the control output of fact;Recycle asynchronous brake unit and controlled quentity controlled variable is acted on controlled device, complete braking procedure.
Description
Technical field
The present invention relates to railway locomotive control field, particularly to a kind of heavy-load combined train based on Model Predictive Control
Asynchronous braking device and method.
Background technology
The current heavy-load combined train of China is mainly started on Datong-Qinhuangdao Railway, uses synchronous braking control system.Due to
No matter train operation is under which kind of orbital environment for it, is all to use consistent braking target value that each locomotive is braked control,
The when of so needed for car body, brake force is inconsistent before and after under train operation is in MODEL OVER COMPLEX TOPOGRAPHY so that car body intermolecular forces
Excessive, even cause the accident such as derailing, disconnected hook beyond its safety range, it is difficult to ensure control for brake performance;Simultaneously because braking
The propagation delay time of instruction will cause train marshalling list unsuitable long.
The train pull weight that is mainly characterized by of heavy haul transport strengthens, and marshalling lengthens, it is achieved omnidistance transit, makes one
Bar railway conveying as much as possible wagon flow, gives full play to railway concentration, large, distance, round-the-clock transport advantage, reaches to carry
High railway transport capacity and efficiency, transport fast freight more, reduce the purpose of cost.Owing to heavy haul transport circuit span is big, residing landform
Complexity, ramp is more, and traditional synchronous braking system cannot well meet train demand, and asynchronous control for brake is to solve
The effective scheme of a heavy haul transport difficult problem.
Asynchronous Brake Control is then to combine the information such as the hitch stress between car body and track operation conditions to make again
Dynamic control, can overcome track grade to differ greatly and the problem of braking instruction delay.Therefore, research heavy-load combined train locomotive
Asynchronous Brake Control, is possible not only to accelerate at a high speed, the starting, also to China's locomotive entirety equipment of heavy-load combined train
Lifting with cargo conveyance capacity has great impetus.
Summary of the invention
It is an object of the invention to develop a kind of asynchronous brake unit of heavy-load combined train based on Model Predictive Control, at row
Car braking procedure considers line environment, system status, solves in synchronous braking control model due to landform and time delay shadow
Ring the problem that the workshop active force caused is excessive, thus reduce the safety problem of heavy-load combined train brake system.
In order to realize above-mentioned technical purpose, the technical scheme is that,
A kind of asynchronous brake unit of heavy-load combined train based on Model Predictive Control, fills including master control locomotive control for brake
Putting, described master control locomotive braking control device includes master control locomotive brake module, control module, main control computer car working condition acquiring mould
Block, human-machine interface unit and main control computer car data transport module, described control module communicates to connect master control locomotive brake respectively
Module, main control computer turner condition acquisition module and main control computer car data transport module, described human-machine interface unit and master control locomotive
Working condition acquiring module communicates to connect, and also includes slave controller car brake control, described slave controller car brake control bag
Include slave controller car brake module, asynchronous controlling module, prediction module, slave controller turner condition acquisition module and slave controller car data to pass
Defeated module, described prediction module communicates to connect asynchronous controlling module, slave controller turner condition acquisition module and slave controller car respectively
Data transmission module, described asynchronous controlling module communication connection slave controller car brake control, described slave controller car number
It is communicatively connected to main control computer car data transport module according to transport module.
The described asynchronous brake unit of a kind of based on Model Predictive Control heavy-load combined train, described asynchronous controlling mould
Block includes
For gathering relevant multiple on-off model, carry out the digital quantity input module of logic control for CPU module;
Digital control amount for being exported by CPU is converted to drive the digital output mould of the signal of various electronic valve
Block;
For transferring level to modulated pwm signal, for controlling equalizing reservoir and the charge valve of checking cylinder and air bleeding valve
Pwm signal output module;
For being that digital signal is sent into chamber volume pressure controller and carried out the analog input that processes by analog-signal transitions
Module;
After digital signal for being received by communication interface is processed by CPU, it is changed into analogue signal, carries out putting of signal
The analog output module of big output;With
For receiving various analog quantity, digital quantity, according to program, various input signals are judged, calculate, output control
System has instructed the central authorities of various function and has processed CPU module;
Described digital output module, digital quantity input module, pwm signal output module, central authorities process CPU module,
Analog output module and Analog input mModule are connected by CAN.
A kind of asynchronous braking method of heavy-load combined train based on Model Predictive Control, uses as arbitrary in claim 1-2
Described device, comprises the following steps:
Step one, the work information of Real-time Collection heavy-load combined train car load;
Step 2, master control locomotive gathers train braking situation, it is judged that train operation state also combines self work information, meter
Calculate and generate the control output being suitable for controlling self-operating state, generate braking instruction simultaneously, and braking instruction is sent to other
Slave controller car control equipment;
Step 3, slave controller car, according to train work information, self work information and the control instruction of master control locomotive, is sentenced
Disconnected the need of being braked, and export by calculating the applicable control controlling self-operating state of generation, feed back to main control computer
The control equipment of car.
Described a kind of based on Model Predictive Control the load asynchronous braking method of unit car, in described step one, work
Condition information includes: unit car length, unit car organizational systems, track grade signal, track curvature radius, locomotive speed are believed
Number, load signal, handle position signal.
The described asynchronous braking method of a kind of based on Model Predictive Control heavy-load combined train, in described step 2,
Described locomotive running state includes: braking, reduction of speed, starts, accelerate.
The described asynchronous braking method of a kind of based on Model Predictive Control heavy-load combined train, in described step 3,
Using T-S model based on satisfaction to carry out judging and calculating, described T-S model is according to the historical information of train operation and control
Input prediction processed output in future, constructs optimal performance index according to the deviation of output in future with train reference track and solves generation
New control input.
The described asynchronous braking method of a kind of based on Model Predictive Control heavy-load combined train, obtains from T-S models
Controlled object model describe with following discrete differential equation:
A(z-1) y (k)=B (z-1)u(k-d)+C(z-1)ξ(k)/Δ
Wherein y (k) and u (k) represents input and the output of system respectively, and d is the prediction step of Model Predictive Control;Δ=
1-s-1Represent difference operator;ξ (k) represents system random disturbances noise;A(z-1)、B(z-1) and C (z-1) it is by least square fitting
Backward shift operator s obtained-1Polynomial matrix;
Reference locus is following k the sampling instant value started from t sampling instant, with from current time real output value y
K () describes for initial first order exponential version:
yr(k+i)=y (k)+[s-y (k)] (1-e-iT/τ) (i=1,2 ...)
Wherein, T is the sampling period, and s is the initial value of reference locus, yrK () is the output valve of reference locus, τ is reference
The time constant of track, e-T/τFor softening coefficient;
Use Diophantine equation obtain jth step after export y (k+j) optimum prediction value:
Y (k+j)=GjΔu(k+j-1)+Ax(k)+Fjy(k)+HjΔu(k-1)+Ejω(k+j)
Wherein dominant vector Δ u (k+j-1) of forecast model, Δ u (k-1) is the brake force that locomotive brake device applies,
State vector x (k) includes following information: hitch stress sequence, train running speed sequence, train weight, rail before and after locomotive
The gradient sequence in road, radius of curvature sequence, output y (k+j) be train in the position of future time instance and speed, wherein k is discrete
Change time series;Gj、Fj、Hj、EjFor Diophantine equation coefficient, ω (k+j) is white noise;
The optimality criterion in k moment takes and exports expected value error containing system, and the two of controlling increment weighted value
Minima minJ (k) of secondary type object function:
Wherein, E represents and seeks expected value, qi, rjIt it is quadratic form coefficient;
Using gradient method to solve above formula optimizing index, obtaining optimal control sequence is:
Δ U (k)=(GTQG+R)-1GTQ[yr(k)-HΔu(k)-Fy(k)]
Δ U (k) is the optimum control series solved, and G, H, F are Diophantine equation coefficient matrix, Q and R is secondary
Type coefficient matrix;
The braking matched with vehicle current operating conditions it is calculated by heavy-load combined train Longitudinal Dynamic Model
Power reference curve, i.e. reference locus, then in conjunction with given braking force value, be calculated reference locus value;Finally utilize feedback
The correction value that correction obtains compares with reference locus, and carries out rolling optimization, obtains controlled quentity controlled variable output.
The described asynchronous braking method of a kind of based on Model Predictive Control heavy-load combined train, feedback correction value yp(k)
By model predication value ymK the weighting ω of () and error amount e (k) determines:
yp(k+i)=ym(k+i)+ω e (k) (i=1,2 ...)
Wherein e (k) is by the k moment actual output y (k) including undeterminable external interference ξ (k) and forecast model
Output ym(k), and the dead time error composition of High-speed Electric sky switch valve.
The method have technical effect that, heavy-load combined train introduces a kind of coordination control mechanism, is guaranteeing master
In the case of control locomotive brake, slave controller car can carry out independent brake control according to own actual situation, eliminates train various
The inconsistent safety problem caused of power demand under orographic condition.
The invention will be further described below in conjunction with the accompanying drawings.
Accompanying drawing explanation
Fig. 1 is the population structure block diagram of present system;
Fig. 2 is the overall construction drawing of asynchronous brake unit;
Fig. 3 is PC104 peripheral interface block diagram;
Fig. 4 is Analog input mModule block diagram;
Fig. 5 is A/D conversion hardware circuit diagram;
Fig. 6 is analog output module block diagram;
Fig. 7 is D/A conversion hardware circuit diagram;
Fig. 8 is digital input circuit figure;
Fig. 9 is digital quantity output circuit figure.
Detailed description of the invention
The asynchronous system dynamic control device of the present invention, its control strategy is as follows:
1, use car body fence thought that reference locus is predicted, solve principal and subordinate and control that locomotive brake force demand is inconsistent asks
Topic;
2, T-S modeling method based on satisfaction is utilized to be predicted model;
3, the method combining generalized predictive control realizes asynchronous braking.
The asynchronous system dynamic control device of the present invention, uses PWM to control technology and closed loop control thought, including digital output
Module, digital quantity input module, pwm signal output module, central authorities' process CPU module, analog output module and analog quantity are defeated
Enter module six part.Wherein, described modules is connected by CAN.
Digital quantity input module processes for 110V on-off model is converted to Transistor-Transistor Logic level feeding CPU module, opens
Pass the amount mainly position of size lock, driver Consoles button and TIM trunk interface module input etc.;
Digital output module drives the opening and closing of various electronic valves for Transistor-Transistor Logic level is converted to 110V on-off model,
Switching valve etc. is controlled including trunk module output and EP.;
It is that digital signal sends into chamber volume Stress control that analog signals input processing module is used for analog-signal transitions
Device processes.Analog signals be mainly equalizing reservoir and checking cylinder EP control part by pressure transducer export continuous
4~20mA current signals, the force value of these current signal representative reservoirs, also have the current signal of other reservoir pressure transducers
And the current signal of effusion meter;
Analog signals output processing module, after the digital signal received by communication interface is processed by CPU, changes
For analogue signal, carry out the amplification output of signal;
Transistor-Transistor Logic level is transferred to 24V on-off model by pwm signal output module, for controlling equalizing reservoir and checking cylinder
Charge valve and air bleeding valve, when charge valve obtains electric, reservoir is inflated, and when vent valve obtains electric, reservoir is exitted, thus realizes equilibrium
Reservoir and the control of brake-cylinder pressure;
Central authorities process the CPU that CPU module is brak control unit, and it receives various analog quantitys, digital quantity,
According to program, various input signals being judged, calculated, output control instruction completes the merits such as control, self-inspection, fault diagnosis
Energy.CPU selects PC/104 control module, and operating system selects embedded real-time operating system QNX, and programming software is selected
ISAGRAF。
See Fig. 1, heavy-load combined train based on Model Predictive Control asynchronous control for brake population structure block diagram.Asynchronous system
Dynamic control output is predicted by reference locus, and rolling optimization and T-S model feedback based on satisfaction correction draw.This device
Intend, by car body fence thought, train being divided into several fence, by multiple fence according to locomotive position and orographic condition
Reference locus is predicted obtaining optimum reference locus;Input variable space is divided into c Fuzzy subspaee by T-S model, i.e.
Rule comprises c rule, and for each Fuzzy subspaee, the partial model of system can describe with a linear equation, and
Total output of system is then each local linear shape model output weighted sum, thus obtains a forecast model structure, in actual control
During system, forgetting factor λ in least square is designed as an adjustable parameter, according to the satisfaction on-line control λ's of system
Value, enable the model of identification quicker, stable approach actual condition.
Its reference locus of the asynchronous brake unit of heavy-load combined train is when following k the sampling that t sampling instant starts
Quarter is worth, and describes with the first order exponential version being initial from current time real output value y (k):
yr(k+i)=y (k)+[s-y (k)] (1-e-iT/τ) (i=1,2 ...) (1)
Wherein: T is the sampling period, s is the initial value of reference locus, yrK () is the output valve of reference locus, τ is reference
The time constant of track.e-T/τFor softening coefficient.The following discrete differential equation of controlled object model obtained from T-S models
Describe:
A(z-1) y (k)=B (z-1)u(k-d)+C(z-1)ξ(k)/Δ (2)
Wherein y (k) and u (k) represents input and the output of system respectively;D is the prediction step of Model Predictive Control;Δ=
1-s-1Represent difference operator;ξ (k) represents system random disturbances noise;A(z-1)、B(z-1) and C (z-1) it is by least square fitting
Backward shift operator s obtained-1Polynomial matrix.Re-use the optimum exporting y (k+j) after Diophantine equation obtains jth step
Predictive value:
Y (k+j)=GjΔu(k+j-1)+Fjy(k)+HjΔu(k-1)+Ejω(k+j) (3)
Locomotive uses T-S model based on satisfaction to be predicted controlling, and its model is according to the historical information of train operation
Export future with controlling input prediction, produce new control input according to output in future and the deviation of train reference track.Prediction
Dominant vector △ u (k+j-1) of model, △ u (k-1) is the brake force that locomotive brake device applies, and state vector x (k) includes
Following information: hitch stress sequence, train running speed sequence, train weight, the gradient sequence of track, curvature before and after locomotive
Radii sequence, output y (k+j) be train in the position of future time instance and speed, wherein k is time discretization sequence.G in formulaj、
Fj、Hj、EjFor Diophantine equation coefficient.ω (k+j) is white noise.
The optimality criterion in k moment takes and exports expected value error containing system, and the two of controlling increment weighted value
The minima of secondary type object function:
Wherein, E represents and seeks expected value, qi, rjIt it is quadratic form coefficient.
Using gradient method to solve the optimizing index of claim 10 again, obtaining optimal control sequence is:
Δ U (k)=(GTQG+R)-1GTQ[Yd(K)-HΔu(k)-Fy(k)] (5)
Δ U (k) is the optimum control series solved, and G, H, F are Diophantine equation coefficient matrix, Q and R is secondary
Type coefficient matrix.
The feedback correction value y of Model Predictive ControlpK () is by model predication value ymK the weighting ω of () and error amount e (k) is certainly
Fixed:
yp(k+1)=ym(k+1)+ω e (k) (i=1,2 ...) (6)
Wherein e (k) can be by the k moment actual output y (k) including undeterminable external interference ξ (k) and prediction mould
Type output ym(k), and the dead time error composition of High-speed Electric sky switch valve.
See Fig. 2, the asynchronous system dynamic control device of the present invention, use PWM to control technology and closed loop control thought, including number
Word amount output module, digital quantity input module, pwm signal output module, central authorities process CPU module, analog output module and
Analog input mModule six part.Wherein, described modules is connected by CAN.
Seeing Fig. 3, CPU module is for the control information according to driver's control room and the status information of brakes self
Carry out logical operations and and miscellaneous equipment between communication.In this device, CPU selects PC/104 control module.CPU module
Be mainly designed to the interface of PC/104 with various peripheral components is designed.Communication interface mainly extends CAN, complete
Become the communication between internal module and external control control unit, owing to needs control logical transition, so extending a piece of
CPLD is used for decoding.LED shows and is mainly used to show some system status informations, the fault message etc. found such as self-inspection, and LED shows
Showing it is to extend by a piece of 8255, charactron is common cathode, and PA, PB, PC of 8255 has connect a charactron respectively.
Seeing Fig. 4, analog input one has 16 tunnels, is used for inputting various analog quantity (such as pressure, electric current, voltage etc.),
The instruction sent including the large and small lock of driver.In order to be able to enable a computer to identify, the analog quantity being transmitted through from sensor needs to turn
It is changed to corresponding digital quantity.From sensor Lai 12 tunnel 4~20mA electric current, be converted to 0~5V voltage through overcurrent conversion chip.
4 tunnel 0~10V voltage signals are converted to 0~5V voltage through OP37.Owing on locomotive, electromagnetic environment is complicated, can be to simulation input
Signal produces interference, must also be filtered eliminating interference before therefore analogue signal carries out A/D converter.Wave filter can be
Passive can also be active, uses Order RC filtering in the present system, and experiment proves that the method is effective.Filtering
The most just carry out A/D conversion, locomotive brake control is required more accurate, so ensureing that sampling resolution at least to have 10.For
The most possible meets required precision, uses the A/D chip ADS8344 of 16 herein.The P1 mouth control signal warp of single-chip microcomputer
After crossing optocoupler 6N136, controlling ADS8344 and carry out A/D conversion, analog quantity inputs from AIN0~AIN7, particular hardware circuit such as Fig. 5
Shown in.
Seeing Fig. 6, analog output channel one has 14 tunnels, and it not only completes chamber volume together with analog input
Stress control, goes back and other unit matching completes other function, and uses CAN during remaining module exchange data.14 ways
Word amount, after light-coupled isolation, carries out D/A conversion, then aims at driving external analog control object by XTR110 and OP37 respectively
4~20mA electric currents and 0~10V voltage.The D/A chip that simulation output uses is TLV5614, single-chip microcomputer P1 mouth control signal warp
Crossing optocoupler and control D/A conversion, hardware circuit is as shown in Figure 7
Seeing Fig. 8, digital quantity input is 64 tunnels, after the 110V on-off model in each input channel detecting system, and warp
After crossing resistor network blood pressure lowering, stabilivolt amplitude limit, capacitor filtering, Phototube Coupling, Schmidt trigger, signal is given at CPU
Reason.
Seeing Fig. 9, digital output is 32 tunnels, and in output channel, CPU receives core bus signal, processes output a certain
The switching signal of individual passage, owing to CPU output signal is Transistor-Transistor Logic level, load capacity is relatively low, in order to brak control unit outside
110V DC operation system be connected, and there is enough driving forces, thus use MOSFET as power amplification element,
And utilize CPU send high frequency modulated control signal by pulse isolation transformer coupling control MOSFET break-make, export specified electricity
Stream is 0.9A/110V.Output has short circuit and protects function, and short circuit protection and setting value is 6A.MOSFET protection circuit can be closed rapidly
The disconnected MOSFET gone wrong.
Population structure according to brak control unit and technology requirement, successfully develop asynchronous brak control unit.Then join
According to the technical performance index requirement of synchronous braking control system, it is carried out operation test.Result shows, the braking of the present invention
Control unit, control accuracy meets synchronous braking technology require performance indications (brak control unit is to gas pressure
Control to stablize time error should control at [-0.5Kp+0.5Kp], it is allowed to overshoot, but overshoot must control within 1.5Kp),
I.e. can meet the actuator as asynchronous braking and realize the demand that train tube pressure accurately controls.
Claims (8)
1. the asynchronous brake unit of heavy-load combined train based on Model Predictive Control, fills including master control locomotive control for brake
Putting, described master control locomotive braking control device includes master control locomotive brake module, control module, main control computer car working condition acquiring mould
Block, human-machine interface unit and main control computer car data transport module, described control module communicates to connect master control locomotive brake respectively
Module, main control computer turner condition acquisition module and main control computer car data transport module, described human-machine interface unit and master control locomotive
Working condition acquiring module communicates to connect, it is characterised in that also include slave controller car brake control, described slave controller car braking
Control device and include slave controller car brake module, asynchronous controlling module, prediction module, slave controller turner condition acquisition module and from control
Locomotive data transport module, described prediction module communicates to connect asynchronous controlling module, slave controller turner condition acquisition module respectively
With slave controller car data transport module, described asynchronous controlling module communicates to connect slave controller car brake control, described
Slave controller car data transport module is communicatively connected to main control computer car data transport module.
A kind of asynchronous brake unit of heavy-load combined train based on Model Predictive Control the most according to claim 1, it is special
Levying and be, described asynchronous controlling module includes
For gathering relevant multiple on-off model, carry out the digital quantity input module of logic control for CPU module;
Digital control amount for being exported by CPU is converted to drive the digital output module of the signal of various electronic valve;
For transferring level to modulated pwm signal, for controlling equalizing reservoir and the charge valve of checking cylinder and the PWM of air bleeding valve
Signal output module;
For being that digital signal is sent into chamber volume pressure controller and carried out the Analog input mModule that processes by analog-signal transitions;
After digital signal for being received by communication interface is processed by CPU, being changed into analogue signal, the amplification carrying out signal is defeated
The analog output module gone out;With
For receiving various analog quantity, digital quantity, according to program, various input signals being judged, calculated, output control refers to
The central authorities having made various function process CPU module;
Described digital output module, digital quantity input module, pwm signal output module, central authorities process CPU module, simulation
Amount output module and Analog input mModule are connected by CAN.
3. the asynchronous braking method of heavy-load combined train based on Model Predictive Control, it is characterised in that use right such as to want
Seek the arbitrary described device of 1-2, comprise the following steps:
Step one, the work information of Real-time Collection heavy-load combined train car load;
Step 2, master control locomotive gathers train braking situation, it is judged that train operation state also combines self work information, calculates raw
Become to be suitable for controlling the control output of self-operating state, generate braking instruction simultaneously, and braking instruction is sent to other from
Control locomotive control equipment;
Step 3, slave controller car is according to train work information, self work information and the control instruction of master control locomotive, it is judged that be
No needs are braked, and export by calculating the control generating applicable control self-operating state, feed back to master control locomotive
Control equipment.
A kind of load asynchronous braking method of unit car based on Model Predictive Control the most according to claim 3, its feature
Being, in described step one, work information includes: unit car length, unit car organizational systems, track grade signal,
Track curvature radius, locomotive speed signal, load signal, handle position signal.
A kind of asynchronous braking method of heavy-load combined train based on Model Predictive Control the most according to claim 3, it is special
Levying and be, in described step 2, described locomotive running state includes: braking, reduction of speed, starts, accelerate.
A kind of asynchronous braking method of heavy-load combined train based on Model Predictive Control the most according to claim 3, it is special
Levy and be, in described step 3, use T-S model based on satisfaction to carry out judging and calculating, described T-S model according to
The historical information of train operation and control input prediction output in future, according to the deviation structure of output in future with train reference track
Optimal performance index also solves the control input that generation is new.
A kind of asynchronous braking method of heavy-load combined train based on Model Predictive Control the most according to claim 6, it is special
Levying and be, the controlled object model obtained from T-S models describes with following discrete differential equation:
A(z-1) y (k)=B (z-1)u(k-d)+C(z-1)ξ(k)/Δ
Wherein y (k) and u (k) represents input and the output of system respectively, and d is the prediction step of Model Predictive Control;Δ=1-s-1
Represent difference operator;ξ (k) represents system random disturbances noise;A(z-1)、B(z-1) and C (z-1) for be obtained by least square fitting
Backward shift operator s-1Polynomial matrix;
Reference locus is following k the sampling instant value started from t sampling instant, with from current time real output value y (k) is
Initial first order exponential version describes:
yr(k+i)=y (k)+[s-y (k)] (1-e-iT/τ) (i=1,2 ...)
Wherein, T is the sampling period, and s is the initial value of reference locus, yrK () is the output valve of reference locus, τ is reference locus
Time constant, e-T/τFor softening coefficient;
Use Diophantine equation obtain jth step after export y (k+j) optimum prediction value:
Y (k+j)=GjΔu(k+j-1)+Ax(k)+Fjy(k)+HjΔu(k-1)+Ejω(k+j)
Wherein dominant vector Δ u (k+j-1) of forecast model, Δ u (k-1) is the brake force that locomotive brake device applies, state
Vector x (k) includes following information: hitch stress sequence before and after locomotive, train running speed sequence, train weight, track
Gradient sequence, radius of curvature sequence, output y (k+j) be train in the position of future time instance and speed, when wherein k is discretization
Between sequence;Gj、Fj、Hj、EjFor Diophantine equation coefficient, ω (k+j) is white noise;
The optimality criterion in k moment takes and exports expected value error containing system, and the quadratic form of controlling increment weighted value
Minima minJ (k) of object function:
Wherein, E represents and seeks expected value, qi, rjIt it is quadratic form coefficient;
Using gradient method to solve above formula optimizing index, obtaining optimal control sequence is:
Δ U (k)=(GTQG+R)-1GTQ[yr(k)-HΔu(k)-Fy(k)]
Δ U (k) is the optimum control series solved, and G, H, F are Diophantine equation coefficient matrix, Q and R is quadratic form system
Matrix number;
The braking Radix Talini Paniculati matched with vehicle current operating conditions it is calculated by heavy-load combined train Longitudinal Dynamic Model
Examine curve, i.e. reference locus, then in conjunction with given braking force value, be calculated reference locus value;Finally utilize feedback compensation
The correction value obtained compares with reference locus, and carries out rolling optimization, obtains controlled quentity controlled variable output.
A kind of asynchronous braking method of heavy-load combined train based on Model Predictive Control the most according to claim 7, it is special
Levy and be, feedback correction value ypK () is by model predication value ymK the weighting ω of () and error amount e (k) determines:
yp(k+i)=ym(k+i)+ω e (k) (i=1,2 ...)
Wherein e (k) is exported by the k moment actual output y (k) including undeterminable external interference ξ (k) and forecast model
ym(k), and the dead time error composition of High-speed Electric sky switch valve.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610853319.8A CN106154840A (en) | 2016-09-27 | 2016-09-27 | A kind of asynchronous braking device and method of heavy-load combined train based on Model Predictive Control |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610853319.8A CN106154840A (en) | 2016-09-27 | 2016-09-27 | A kind of asynchronous braking device and method of heavy-load combined train based on Model Predictive Control |
Publications (1)
Publication Number | Publication Date |
---|---|
CN106154840A true CN106154840A (en) | 2016-11-23 |
Family
ID=57340999
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610853319.8A Pending CN106154840A (en) | 2016-09-27 | 2016-09-27 | A kind of asynchronous braking device and method of heavy-load combined train based on Model Predictive Control |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106154840A (en) |
Cited By (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107640183A (en) * | 2017-07-31 | 2018-01-30 | 李振轩 | A kind of operation control method for train based on iterative learning control |
CN109263680A (en) * | 2018-09-11 | 2019-01-25 | 河南机电职业学院 | A kind of municipal rail train braking control system based on Fuzzy Predictive Control |
CN109598088A (en) * | 2018-12-20 | 2019-04-09 | 中国矿业大学 | A kind of belt conveyor belt speed setting control method |
CN110450794A (en) * | 2019-08-26 | 2019-11-15 | 西南交通大学 | A kind of optimized adhesion control method based on optimal creep speed search and tracking |
CN113703316A (en) * | 2021-07-23 | 2021-11-26 | 上海电力大学 | Prediction control method for main steam pressure model of pressurized water reactor nuclear power station |
CN113997976A (en) * | 2020-07-28 | 2022-02-01 | 株洲中车时代电气股份有限公司 | Asynchronous control method and related device for heavy-load combined train |
CN114633780A (en) * | 2021-12-30 | 2022-06-17 | 中南大学 | Heavy haul train and longitudinal dynamics traction operation optimization control system thereof |
CN118034067A (en) * | 2024-04-11 | 2024-05-14 | 保定博堃元信息科技有限公司 | Industrial process model parameter optimization method based on generalized predictive control |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101497343A (en) * | 2009-01-22 | 2009-08-05 | 北京全路通信信号研究设计院 | Operation control method and operation control system of combined train |
CN101537841A (en) * | 2009-04-22 | 2009-09-23 | 北京人和路通科技有限公司 | Control method of a heavy load combined train marshaled mixedly by locomotives and system thereof |
CN203465763U (en) * | 2013-09-30 | 2014-03-05 | 长沙瑞纬电气有限公司 | Heavy-load combined train brake valve cut-out recording analysis device |
JP5450121B2 (en) * | 2010-01-15 | 2014-03-26 | 太平洋セメント株式会社 | Model predictive control device and program |
-
2016
- 2016-09-27 CN CN201610853319.8A patent/CN106154840A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101497343A (en) * | 2009-01-22 | 2009-08-05 | 北京全路通信信号研究设计院 | Operation control method and operation control system of combined train |
CN101537841A (en) * | 2009-04-22 | 2009-09-23 | 北京人和路通科技有限公司 | Control method of a heavy load combined train marshaled mixedly by locomotives and system thereof |
JP5450121B2 (en) * | 2010-01-15 | 2014-03-26 | 太平洋セメント株式会社 | Model predictive control device and program |
CN203465763U (en) * | 2013-09-30 | 2014-03-05 | 长沙瑞纬电气有限公司 | Heavy-load combined train brake valve cut-out recording analysis device |
Non-Patent Citations (2)
Title |
---|
刘剑锋: "基于模糊模型预测控制的重载组合列车机车制动控制策略研究", 《中国博士学位论文全文数据库 信息科技辑》 * |
童海涛: "重载组合列车异步制动控制技术的研究", 《中国优秀硕士学位论文全文数据库 工程科技II辑》 * |
Cited By (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107640183A (en) * | 2017-07-31 | 2018-01-30 | 李振轩 | A kind of operation control method for train based on iterative learning control |
CN109263680A (en) * | 2018-09-11 | 2019-01-25 | 河南机电职业学院 | A kind of municipal rail train braking control system based on Fuzzy Predictive Control |
CN109598088A (en) * | 2018-12-20 | 2019-04-09 | 中国矿业大学 | A kind of belt conveyor belt speed setting control method |
CN109598088B (en) * | 2018-12-20 | 2022-02-11 | 中国矿业大学 | Belt speed setting control method for belt conveyor |
CN110450794A (en) * | 2019-08-26 | 2019-11-15 | 西南交通大学 | A kind of optimized adhesion control method based on optimal creep speed search and tracking |
CN113997976A (en) * | 2020-07-28 | 2022-02-01 | 株洲中车时代电气股份有限公司 | Asynchronous control method and related device for heavy-load combined train |
CN113997976B (en) * | 2020-07-28 | 2024-01-23 | 株洲中车时代电气股份有限公司 | Asynchronous control method and related device for heavy-duty combined train |
CN113703316A (en) * | 2021-07-23 | 2021-11-26 | 上海电力大学 | Prediction control method for main steam pressure model of pressurized water reactor nuclear power station |
CN114633780A (en) * | 2021-12-30 | 2022-06-17 | 中南大学 | Heavy haul train and longitudinal dynamics traction operation optimization control system thereof |
CN114633780B (en) * | 2021-12-30 | 2022-11-25 | 中南大学 | Heavy-duty train and longitudinal dynamics traction operation optimization control system thereof |
CN118034067A (en) * | 2024-04-11 | 2024-05-14 | 保定博堃元信息科技有限公司 | Industrial process model parameter optimization method based on generalized predictive control |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106154840A (en) | A kind of asynchronous braking device and method of heavy-load combined train based on Model Predictive Control | |
EP3628529B1 (en) | Train compartment brake control method, train compartment, and train | |
CN104590333A (en) | Railway train intelligent operation optimization control system | |
EP1984217B1 (en) | Distributed train intelligence system and method | |
CN103863364B (en) | A kind of freight locomotive based on scheduling signals handles real-time optimal control system automatically | |
Li et al. | Neural adaptive fault tolerant control for high speed trains considering actuation notches and antiskid constraints | |
CN104627201B (en) | Microcomputer controlled electro sky brake unit for locomotive braking system | |
CN109572654B (en) | Impact rate control method based on traction braking fusion control system | |
CN107943022A (en) | A kind of PID locomotive automatic Pilot optimal control methods based on intensified learning | |
Bai et al. | Energy-efficient locomotive operation for Chinese mainline railways by fuzzy predictive control | |
CN109693688B (en) | Automatic train driving control system with preset performance and control method | |
CN105365851A (en) | Intelligent trainman control device | |
CN104169908B (en) | Train control system | |
RU2666499C1 (en) | Vehicle operation method | |
Kim et al. | Precise stop control and experimental validation for metro train overcoming delays and nonlinearities | |
Yang et al. | Research on train dynamic coupling strategy based on distributed model predictive control | |
US20210323589A1 (en) | System and method for associating wireless nodes of a consist | |
CN115465332B (en) | Control method, system, electronic equipment and storage medium for motor train unit train | |
Zhou et al. | Model predictive control for high-speed train with automatic trajectory configuration and tractive force optimization | |
RU2577196C1 (en) | Method of transmitting information messages in microprocessor control and diagnostic systems | |
WO2013184026A1 (en) | Microprocessor system for controlling and diagnosing a locomotive with a hybrid drive system | |
Babkov et al. | To the issue of creating a “smart” locomotive | |
RU2758581C1 (en) | Automated system and method for car retarder pneumatic drive control | |
Guo et al. | Distributed Adaptive Fault-Tolerant Control for High-Speed Trains using Multi-Agent System Model | |
Pan et al. | Design of multi-objective optimal control algorithm for automatic driving of EMU based on PSO |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20161123 |