CN106647280A - HEV (hybrid electric vehicle) mode switching data driving prediction controller design method and HEV mode switching data driving prediction controller design system - Google Patents
HEV (hybrid electric vehicle) mode switching data driving prediction controller design method and HEV mode switching data driving prediction controller design system Download PDFInfo
- Publication number
- CN106647280A CN106647280A CN201710028468.5A CN201710028468A CN106647280A CN 106647280 A CN106647280 A CN 106647280A CN 201710028468 A CN201710028468 A CN 201710028468A CN 106647280 A CN106647280 A CN 106647280A
- Authority
- CN
- China
- Prior art keywords
- data
- hybrid vehicle
- mode switching
- module
- predictive
- 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.)
- Granted
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)
- Hybrid Electric Vehicles (AREA)
- Electric Propulsion And Braking For Vehicles (AREA)
Abstract
The invention discloses an HEV mode switching data driving prediction controller design method. The HEV mode switching data driving prediction controller design method comprises steps that a whole hybrid electric vehicle simulation model is established in vehicle simulation software; motor torque, engine torque, and clutch torque are designed as excitation data, and the designed excitation data is used for the established hybrid electric vehicle simulation model; the input and output data block Hankel matrix of the hybrid electric vehicle is established based on the acquired open loop data, and then a prediction output equation of a hybrid electric vehicle mode switching system is established based on an identification model; an increment of an input sequence, an output sequence, and a future input sequence of a data driving prediction controller are defined, and the prediction output equation is expressed in an increment way; the data driving prediction controller is solved in an on-line manner according to a corresponding input constraint and a corresponding output constraint. Calculation amount is reduced, and a modeling error problem caused by a switching process from the data identification model to a state space model is prevented.
Description
Technical field
The invention belongs to field of hybrid electric vehicles, more particularly to a kind of data-driven PREDICTIVE CONTROL of HEV mode switching
Device method for designing and its system.
Background technology
The English full name of HEV is Hybrid Electric Vehicle, and Chinese is：Hybrid vehicle.
With energy shortage and the aggravation of problem of environmental pollution, hybrid vehicle (Hybrid Electric
Vehicle, HEV) the fuel consumption and emission performance of vehicle can be greatly improved, enough continual mileages are can guarantee that again, be
A kind of feasible program for reducing Fossil fuel consumption and carbon emission effective in the near future, it has also become current study hotspot and main flow
Direction.
Current some advanced theories, such as：Fuzzy self-adaption sliding-mode method, model matching control, model reference control, cut
The methods such as control, Model Predictive Control are changed, this domain variability is applied to and is achieved many achievements in research, but these methods are all
Based on mechanism model, its control effect depends critically upon the accuracy of modeling.And HEV power assembly system complex structures, and
And system performance and each parameters of operating part will change (for example, the damping system of axle with the long-term ageing of HEV and the change of driving conditions
Great changes have taken place for several changes meetings with ambient temperature), therefore, it is difficult to setting up accurate mechanism model to it.Enable and set up description
The global mathematical models of HEV power assembly systems, surely in high-order nonlinear.High-order nonlinear model is difficult with based on mechanism
The control method of model is processed.I.e. enable is processed, and the controller for also necessarily causing high-order complicated, operand is quite huge,
The rapidity and requirement of real-time of HEV can not be met.In view of this, existing research has all been done at certain simplification in modeling
Reason, but certainly will there is many Unmarried pregnancies, cause control effect not good.
The content of the invention
In order to solve the shortcoming of prior art, the first object of the present invention is to provide a kind of data of HEV mode switching and drives
Dynamic Design of Predictive method.
A kind of data-driven Design of Predictive method of the HEV mode switching of the present invention, including：
Step 1：Whole hybrid vehicle phantom is built in automobile simulation software；
Step 2：According to the dynamic characteristic of hybrid vehicle mode switching system, design motor torque, motor torque
With clutch torque as excited data, and designed excited data is acted on into built hybrid vehicle emulation mould
Type；
Step 3：Data block Hankel of input and the output of hybrid vehicle is built based on the open loop data of collection
Matrix, and then the prediction output equation of hybrid vehicle mode switching system is constructed based on identification model；
Step 4：Define hybrid vehicle mode switching system the increment of list entries, output sequence and future it is defeated
Enter sequence, prediction output equation is stated with incremental form；
Step 5：According to corresponding input constraint and output constraint, during building the pure electronic switching to hybrid mode
The cost function of optimum control, and the prediction output equation based on incremental form statement is come line solver data-driven PREDICTIVE CONTROL
Device.
The hybrid vehicle phantom includes engine block, motor module, clutch module, change speed gear box mould
Block, battery module, tire module, differential module and longitudinal direction of car kinetics.
The hybrid vehicle phantom of the present invention is the whole vehicle model that clutch enables single shaft parallel connection HEV, and the HEV is whole
Vehicle model can well portray the Transient Dynamics characteristic of vehicle, for example, the lag characteristic of engine output torque, clutch
Torsional shaking, the vibration characteristics of drive shaft and the slip rate characteristic of tire etc..The HEV includes parking, pure electronic, combined drive
Dynamic, pure engine driving, braking energy feedback, charging multiple-working mode.
Wherein, the data-driven forecast Control Algorithm of data-driven predictive controller is by Subspace Identification and model prediction control
Make the composition that combines.The present invention combines to realize data-driven PREDICTIVE CONTROL using Subspace Identification and Model Predictive Control
Method, eliminates in the method based on mechanism model to the solution of state space equation, has both reduced amount of calculation, turn avoid from
The modeling error problem introduced when data identification is to state-space model.
In the step 3, according to the prediction output equation of hybrid vehicle mode switching system, distinguished using subspace
Knowledge method is estimating the following output valve of hybrid vehicle mode switching system.
The method, using simple, ensure that the robustness of numerical value without the need for advance model parameterization.
Cost function is made up of two parts, wherein, the Part I of cost function forces clutch slip speed convergence to arrive
0, so as to realizing rapidly pattern switching and reducing clutch abrasion；Cost function Part II control motor torque, from
Clutch torque and the rate of change of motor torque, for the comfortableness of Assured Mode switching.
The second object of the present invention is to provide a kind of data-driven Design of Predictive system of HEV mode switching.
A kind of data-driven Design of Predictive system of the HEV mode switching of the present invention, including：
Hybrid vehicle phantom builds module, and it is used to build whole hybrid vehicle phantom；
Hybrid vehicle mode switching system builds module, and it is used for according to hybrid vehicle mode switching system
Dynamic characteristic, design motor torque, motor torque and clutch torque as excited data, and by designed excited data
Act on built hybrid vehicle phantom；
Prediction output equation constructing module, it is used to build the input of hybrid vehicle based on the open loop data of collection
With the data block Hankel matrix of output, and then the prediction of hybrid vehicle mode switching system is constructed based on identification model
Output equation；
Incremental form statement prediction output equation module, it is used to define the input of hybrid vehicle mode switching system
The list entries of the increment, output sequence and future of sequence, prediction output equation is stated with incremental form；
Data-driven predictive controller solves module, and it is used for according to corresponding input constraint and output constraint, builds pure
The cost function of the optimum control during the electronic switching to hybrid mode, and the prediction output side stated based on incremental form
Journey carrys out line solver data-driven predictive controller.
The hybrid vehicle phantom include engine block, motor module, clutch module, battery module,
Tire module, speed change tank module, differential module and longitudinal direction of car kinetics.The hybrid vehicle phantom of the present invention
For the whole vehicle model that clutch enables single shaft parallel connection HEV, the HEV whole vehicle models can well portray the Transient Dynamics of vehicle
Characteristic, for example, the lag characteristic of engine output torque, the torsional shaking of clutch, the vibration characteristics and tire of drive shaft
Slip rate characteristic etc..The HEV includes that parking, pure electronic, combination drive, pure engine driving, braking energy feedback, charging are more
Plant mode of operation.
Build in module in hybrid vehicle mode switching system, the data-driven prediction of data-driven predictive controller
Control method combines what is constituted by Subspace Identification and Model Predictive Control.
The present invention combines to realize data-driven forecast Control Algorithm using Subspace Identification and Model Predictive Control, saves
Go in the method based on mechanism model to the solution of state space equation, both reduced amount of calculation, turn avoid from data and distinguish
Know the modeling error problem introduced during state-space model.
In the prediction output equation constructing module, according to the prediction output side of hybrid vehicle mode switching system
Journey, the following output valve of hybrid vehicle mode switching system is estimated using subspace state space system identification.The method is without the need for pre-
First model parameterization, using simple, ensure that the robustness of numerical value.
Cost function is made up of two parts, wherein, the Part I of cost function forces clutch slip speed convergence to arrive
0, so as to realizing rapidly pattern switching and reducing clutch abrasion；Cost function Part II control motor torque, from
Clutch torque and the rate of change of motor torque, for the comfortableness of Assured Mode switching.
Beneficial effects of the present invention are：
(1) the data-driven Design of Predictive method of a kind of HEV mode switching according to the present invention, wherein, data
The control method of predictive controller is driven, traditional controlling party based on mechanism model is substituted using the Forecasting Methodology based on data
Method, overcomes real-time with control for the accuracy of the modelling by mechanism in the presence of complication system as HEV power assemblies
Contradiction between property.Compared with traditional model predictive control method, the data that method for designing of the present invention is designed
The pattern switching quality of driving predictive controller is higher, and this is favorably improved the ride comfort of HEV, while extending clutch
Service life.
(2) torque coordination being used in HEV work-mode switching process using advanced data-driven forecast Control Algorithm
In control, it is directly based upon the offline inputoutput data for gathering to design controller, as effectively prevent HEV power assemblies
The problem of the modelling by mechanism hardly possible in the presence of complication system, eliminates in the method based on mechanism model to state space equation
Solve, both reduced amount of calculation, the modeling error problem introduced when turn avoid from data identification to state-space model.
(3) control algolithm of the present invention only needs to do some off-line simulations experiments and is not required to obtaining inputoutput data
Any experience and rating test are wanted, development cost is low.As time goes on, system performance and parameters of operating part may be total with power
Gradually change into the aging of part, only need vehicle operation data again needed for collection algorithm, and without the need for algorithm
Do any change, you can persistently ensure the excellent control performance of institute's extracting method.
Description of the drawings
Fig. 1 is whole hybrid vehicle phantom the general frame；
Fig. 2 is data-driven predictive control algorithm block diagram；
Fig. 3 (a) is the input for recognizing the 219 of forecast model sampled points and output data --- motor torque
Te；
Fig. 3 (b) is the input for recognizing the 219 of forecast model sampled points and output data --- clutch torque
Tc；
Fig. 3 (c) is the input for recognizing the 219 of forecast model sampled points and output data --- motor torque Tm；
Fig. 3 (d) is the input for recognizing the 219 of forecast model sampled points and output data --- clutch both sides
Speed difference Δ ω；
Fig. 3 (e) is the input for recognizing the 219 of forecast model sampled points and output data --- engine speed
ωe；
Fig. 3 (f) is the input for recognizing the 219 of forecast model sampled points and output data --- motor speed ωm；
Fig. 4 (a) is the input data-motor torque T for checkinge；
Fig. 4 (b) is input data --- the clutch torque T for checkingc；
Fig. 4 (c) is input data --- the motor torque T for checkingm；
Fig. 4 (d) is the identification result for exporting Δ ω；
Fig. 4 (e) is constraint output ωeIdentification result；
Fig. 4 (f) is constraint output ωmIdentification result；
Fig. 5 is to realize the Simulink block figures based on the torque coordination control strategy of data-driven forecast Control Algorithm；
Fig. 6 (a) is the speed of UDC operating modes；
Fig. 6 (b) is the vehicle impact degree controlled using the torque coordination based on data-driven forecast Control Algorithm；
Fig. 6 (c) is the speed of the clutch both sides controlled using the torque coordination based on data-driven forecast Control Algorithm
Difference；
Fig. 6 (d) is the motor torque controlled using the torque coordination based on data-driven forecast Control Algorithm；
Fig. 6 (e) is the clutch torque controlled using the torque coordination based on data-driven forecast Control Algorithm；
Fig. 6 (f) is the motor torque controlled using the torque coordination based on data-driven forecast Control Algorithm；
Fig. 7 (a) is the comparison of data-driven forecast Control Algorithm and model predictive control method --- clutch both sides
Speed difference；
Fig. 7 (b) is the partial enlargement of Fig. 7 (a)；
Fig. 7 (c) is the comparison of data-driven forecast Control Algorithm and model predictive control method --- shock extent；
Fig. 7 (d) is the comparison of data-driven forecast Control Algorithm and model predictive control method --- motor speed and send out
Motivation rotating speed；
Fig. 7 (e) is the partial enlargement of Fig. 7 (d).
Specific embodiment
Below in conjunction with the accompanying drawings the present invention will be further described with embodiment：
(1) Fig. 1 is the whole hybrid vehicle phantom the general frame of the present invention.In professional automobile simulation software
Certain clutch built in Cruise enables the whole vehicle model of single shaft parallel connection HEV.Whole hybrid vehicle phantom is main
Including engine block, motor module, clutch module, battery module, tire module, speed change tank module, differential module with
And longitudinal direction of car kinetics, whole vehicle model parameter such as table 1.
The hybrid vehicle parameter list of table 1
Constructed HEV whole vehicle models can well portray the Transient Dynamics characteristic of vehicle, for example, electromotor output
The lag characteristic of torque, the torsional shaking of clutch, the vibration characteristics of drive shaft and the slip rate characteristic of tire etc..Electromotor
Power is delivered to drivetrain by the friction torque of clutch.Management level controller uses rule-based energy management plan
Slightly.In order to improve fuel economy, reduce discharge, the HEV includes parking, pure electronic, combination drive, pure engine driving, system
The multiple-working modes such as energy feedback, charging.
(2) the HEV mode of operation switching torque coordination control algolithms in the present invention use the pre- observing and controlling of data-driven
Method processed, is combined by Subspace Identification and Model Predictive Control and is constituted, and it is as shown in Figure 2 that it constitutes schematic diagram.Specifically
Embodiment is as follows：
1. open loop data collection：It is designed to reach and continues fully to encourage electromotor, clutch, motor, drive shaft and wheel
The input T of the vibration characteristics of tiree、Tc、Tm, as shown in Fig. 3 (a) -3 (c) input datas, acted on built HEV car loads
On kinetic model, system controlled output Δ ω and affined output ω is obtainede、ωm, such as shown in Fig. 3 (d) -3 (f).This
Bright middle sampling time Ts=0.01s, i=20, j=180, therefore altogether collection electronic switches dynamic mistake from pure to hybrid mode
The open loop data of 2i+j-1=219 sampling instant in journey.
Open loop data includes：The speed difference of clutch both sides, engine speed and motor speed.
2. based on the sampled data shown in Fig. 3 (a) -3 (f), the data block Hankel matrix U of u (k) and y (k) is builtp、Uf、
Yp、Yf、Subscript p and f of matrix is represented over respectively and future.
3. input and output Hankel matrix Us are passed throughp、Uf、Yp、Yf、The prediction output equation for constructing system is：
Wherein
It is the system future output valve estimated by subspace state space system identification, using least square method, solving-optimizing
Problem (3), obtains two sub-spaces matrix Ls in predictive equationw,Lu
Can try to achieve in the same manner
4. with the 220th sampling instant to the inputoutput data of the 599th sampling instant verifying that it is pre- that identification is obtained
Survey equation.Constructed input data T of excitation completelye、Tc、TmAs shown in Fig. 4 (a) -4 (c), specific identification prediction effect is such as
Fig. 4 (d) -4 (f).As shown in Fig. 4 (d) -4 (f), when the speed difference of clutch both sides is not 0, the output for obtaining is predicted
Can coincide well with real output；It is near 0, because identification prediction that forecast error occurs in clutch both sides speed difference
Sampled data during model is not including the situation that speed difference is 0.What the present invention was studied is that the cunning of clutch is rubbed the stage, in this stage
It is not 0 that clutch both sides speed difference is, therefore does not affect the design of data-driven predictive controller.
5. the increment for defining the control input sequence at k moment to be optimized is：
Wherein,
Define PREDICTIVE CONTROL output sequenceWith the list entries u in futurefK () is as follows：
NpAnd NuPrediction time domain and control time domain are represented respectively.In order that prediction is exported close to given with reference to defeated
Go out, generally cost function is defined as into quadric form, in order to punish the acute variation of controlled quentity controlled variable, introduce the punishment to controlled quentity controlled variable
, accordingly, it is considered to input constraint and affined output constraint are arrived, the optimum control during the pure electronic switching to hybrid mode
Problem processed can be described as follows：
s.t.umin(k+q)≤u(k+q)≤umax(k+q), q=0,1 ..., Nu-1 (5.2)
Δumin(k+q)≤Δu(k+q)≤Δu max(k+q), q=0,1 ..., Nu-1 (5.3)
ybmin(k+q)≤yb(k+q)≤ybmax(k+q), q=0,1 ..., Np (5.4)
Wherein
The Part I of cost function J forces clutch slip speed convergence to 0, so as to realize quick pattern switching and
Reduce clutch abrasion；The rate of change of the Part II control motor torque, clutch torque and motor torque of cost function,
The comfortableness of Assured Mode switching.Re(k+1) be prediction time domain in Δ ω reference sequences, wherein α ∈ (0,1) be one can
The parameter of regulation.Because the speed difference of the main slave end of clutch when y (k) represents current time k, is gradually to level off to 0, institute
It is less with α, then reference value r (k+i) of y (k+i), i=1,2 ..., NpIt is less, so as to realize the speed of the main slave end of clutch
Difference reaches as early as possible 0, so as to complete the combination of clutch sooner.
6. in order to solve optimal problem (5.1), prediction output side is derived based on data-driven method and forecast Control Algorithm
Journey.In order that system output can be input into zero steady-state error track reference, subspace predictive equation is expressed as into following increasing
Amount form：
Wherein
The controlled output increment sequence that prediction is obtained is represented by：
Wherein Δ yp=[Δ y (k-i+1) Δ y (k-i+2) ... Δs y (k)]T, Δ up=[Δ u (k-i) Δ u (k-i+
1) … Δu(k-1)]T。
The optimum prediction that following output can be obtained is as follows
WhereinIt is Npm×NuL dynamic matrixs.
Y (k)=[y (k) y (k) ... y (k)]T (10)
According to (9)-(12), can be derived from the same manner withRelevant
7. the torque coordination control problem (5.1) for considering constraint is converted into into following quadratic programming problem, so as to ask online
Solution data-driven predictive controller.
s.t.CuΔuf(k)≤b
Wherein
According to the roll stablized loop principle of PREDICTIVE CONTROL, only by first element interaction of Δ u (k) in HEV drivetrains
System, at each moment above-mentioned calculating is repeated.Therefore, it is as follows in the control law of k-th sampling instant：
Above-mentioned data-driven predictive control algorithm is realized in Matlab/Simulink, as shown in Figure 5.
(3) in order to verify the control performance of torque coordination data-driven predictive controller, in the urban district as shown in Fig. 6 (a)
Under driving cycles UDC (urban driving cycle), the pre- observing and controlling of torque coordination data-driven that will be realized in Simulink
The HEV Full Vehicle Dynamics models built in device processed and Cruise, by api interface mode associative simulation is realized.In operation
During Cruise phantoms, Simulink phantoms are also calculated at the same time and solved, in simulation process between the two not
The disconnected exchange for carrying out data.Cruise simulation step lengths are 0.01 second.Electromotor, clutch, the peak torque such as institute of table 1 of motor
Show.Np=20, Nu=5, ε=30rad/s, ωemin=720r/min, ωmmin=0r/min, α=0.8, τy,1=τy,2=...=
τy,20=0.8,In order to improve the ride comfort of clutch engagement to improve driving
Comfortableness, the excursion of controlling increment is set to：-50Nm/s≤ΔTe≤ 50Nm/s, -50Nm/s≤Δ Tc≤50Nm/
S, -100Nm/s≤Δ Tm≤100Nm/s。
First, the control performance of torque coordination data-driven predictive controller itself is demonstrated.Simulation result such as Fig. 6 (b)-
Shown in 6 (f).Start within 60.44th second speed more than 27km/h, electromotor is started with the Start-up of constant torque of 60Nm when the 61.22nd second
Machine rotating speed rises to close motor speed (differing 30rad/s with motor speed), and clutch initially enters the sliding friction stage, the
The speed difference of clutch both sides is 0 when 61.49 seconds, according to clutch characteristics, makes clutch torque in a kind of predefined mode
Increase, to ensure its secure ratcs.From simulation result, put forward the pattern switching quality is good of strategy, this from it is pure it is electronic to
The mode handover procedure of hybrid power is needed altogether 1.05 seconds, can well meet the switching demand of driver, vehicle impact degree width
It is worth for 9.96m/s3, it is to avoid discomfort of the passenger in mode handover procedure, the clutch slipping time is only 0.27 second, it is total from
Clutch unit friction work is only 8.59J, and this contributes to extending clutch service life.
Secondly, the control performance of torque coordination data-driven predictive controller and traditional model predictive control method are entered
Go and compared.Shown in simulation result such as Fig. 7 (a) -7 (e).In order to more intuitively be compared to two methods, table 2 lists energy
The several important simulation result of enough representation pattern quality of handoff, it is seen then that under identical simulated conditions, is predicted based on data-driven
The shock extent of the mode handover procedure of control method is less, and the pattern switching persistent period is shorter, and the clutch slipping time is shorter, from
It is less that clutch slides abrasion consumption.
The data-driven forecast Control Algorithm of table 2 compares with the simulation result of model predictive control method
| Data-driven forecast Control Algorithm | Model predictive control method | |
| Shock extent fluctuation range | - 9.96～6.76m/s3 | - 82.47～87.92m/s3 |
| The pattern switching persistent period | 1.05s | 1.3s |
| The clutch slipping time | 0.27s | 0.5s |
| Clutch slipping is lost | 8.59J | 19.3J |
The present invention relates to a kind of HEV mode of operations switching torque coordination control method, using the prediction side based on data
Method substitutes the traditional control method based on mechanism model, overcomes for being deposited in complication system as HEV power assemblies
The accuracy of modelling by mechanism and the real-time of control between contradiction.Compared with traditional model predictive control method, this
The adopted method mode quality of handoff of invention is higher, and this is favorably improved the ride comfort of HEV, while extending clutch
Service life.
Although the above-mentioned accompanying drawing that combines is described to the specific embodiment of the present invention, not to present invention protection model
The restriction enclosed, one of ordinary skill in the art should be understood that on the basis of technical scheme those skilled in the art are not
Need the various modifications made by paying creative work or deformation still within protection scope of the present invention.
Claims (10)
1. a kind of data-driven Design of Predictive method that HEV mode switches, it is characterised in that include：
Step 1：Whole hybrid vehicle phantom is built in automobile simulation software；
Step 2：According to the dynamic characteristic of hybrid vehicle mode switching system, design motor torque, motor torque and from
Designed excited data is acted on built hybrid vehicle phantom by clutch torque as excited data；
Step 3：The data block Hankel matrix of input and the output of hybrid vehicle is built based on the open loop data of collection,
The prediction output equation of hybrid vehicle mode switching system is further constructed based on identification model；
Step 4：Define the input sequence of the increment, output sequence and future of the list entries of hybrid vehicle mode switching system
Row, prediction output equation is stated with incremental form；
Step 5：According to corresponding input constraint and output constraint, the optimum during the pure electronic switching to hybrid mode is built
The cost function of control, and the prediction output equation based on incremental form statement is come line solver data-driven predictive controller.
2. the data-driven Design of Predictive method that a kind of HEV mode as claimed in claim 1 switches, its feature exists
In the hybrid vehicle phantom includes engine block, motor module, clutch module, speed change tank module, battery
Module, tire module, differential module and longitudinal direction of car kinetics.
3. the data-driven Design of Predictive method that a kind of HEV mode as claimed in claim 1 switches, its feature exists
Combined structure by Subspace Identification and Model Predictive Control in the data-driven forecast Control Algorithm of, data-driven predictive controller
Into.
4. the data-driven Design of Predictive method that a kind of HEV mode as claimed in claim 1 switches, its feature exists
In in the step 3, according to the prediction output equation of hybrid vehicle mode switching system, using Subspace Identification side
Method is estimating the following output valve of hybrid vehicle mode switching system.
5. the data-driven Design of Predictive method that a kind of HEV mode as claimed in claim 1 switches, its feature exists
In, cost function is made up of two parts, wherein, the Part I of cost function forces clutch slip speed convergence to 0, so as to
Realize rapidly pattern switching and reduce clutch abrasion；The Part II control motor torque of cost function, clutch turn
The rate of change of square and motor torque, for the comfortableness of Assured Mode switching.
6. the data-driven Design of Predictive system that a kind of HEV mode switches, it is characterised in that include：
Hybrid vehicle phantom builds module, and it is used to build whole hybrid vehicle phantom；
Hybrid vehicle mode switching system builds module, and it is used for the dynamic according to hybrid vehicle mode switching system
Characteristic, design motor torque, motor torque and clutch torque act on designed excited data as excited data
In the hybrid vehicle phantom built；
Prediction output equation constructing module, it is used to build the input of hybrid vehicle and defeated based on the open loop data for gathering
The data block Hankel matrix for going out, and then the prediction output of hybrid vehicle mode switching system is constructed based on identification model
Equation；
Incremental form statement prediction output equation module, it is used to define the list entries of hybrid vehicle mode switching system
Increment, output sequence and future list entries, prediction output equation is stated with incremental form；
Data-driven predictive controller solves module, and it is used for according to corresponding input constraint and output constraint, builds pure electronic
To hybrid mode switch during optimum control cost function, and based on incremental form state prediction output equation come
Line solver data-driven predictive controller.
7. the data-driven Design of Predictive system that a kind of HEV mode as claimed in claim 6 switches, its feature exists
In the hybrid vehicle phantom includes engine block, motor module, clutch module, speed change tank module, battery
Module, tire module, differential module and longitudinal direction of car kinetics.
8. the data-driven Design of Predictive system that a kind of HEV mode as claimed in claim 6 switches, its feature exists
In building in module in hybrid vehicle mode switching system, the data-driven PREDICTIVE CONTROL of data-driven predictive controller
Method combines what is constituted by Subspace Identification and Model Predictive Control.
9. the data-driven Design of Predictive system that a kind of HEV mode as claimed in claim 6 switches, its feature exists
In in the prediction output equation constructing module, according to the prediction output equation of hybrid vehicle mode switching system, profit
The following output valve of hybrid vehicle mode switching system is estimated with subspace state space system identification.
10. the data-driven Design of Predictive system that a kind of HEV mode as claimed in claim 6 switches, its feature exists
In, cost function is made up of two parts, wherein, the Part I of cost function forces clutch slip speed convergence to 0, so as to
Realize rapidly pattern switching and reduce clutch abrasion；The Part II control motor torque of cost function, clutch turn
The rate of change of square and motor torque, for the comfortableness of Assured Mode switching.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710028468.5A CN106647280B (en) | 2017-01-16 | 2017-01-16 | The data-driven Design of Predictive method and system of HEV mode switching |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710028468.5A CN106647280B (en) | 2017-01-16 | 2017-01-16 | The data-driven Design of Predictive method and system of HEV mode switching |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106647280A true CN106647280A (en) | 2017-05-10 |
| CN106647280B CN106647280B (en) | 2019-05-24 |
Family
ID=58843331
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710028468.5A Expired - Fee Related CN106647280B (en) | 2017-01-16 | 2017-01-16 | The data-driven Design of Predictive method and system of HEV mode switching |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106647280B (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107168319A (en) * | 2017-06-01 | 2017-09-15 | 同济大学 | A kind of unmanned vehicle barrier-avoiding method based on Model Predictive Control |
| CN108181811A (en) * | 2017-12-18 | 2018-06-19 | 长春工业大学 | A kind of slip rate tracking and controlling method based on linear time-varying |
| CN111301397A (en) * | 2020-02-28 | 2020-06-19 | 南京航空航天大学 | Variable time domain model prediction energy management method for plug-in hybrid electric vehicle |
| CN112318511A (en) * | 2020-11-02 | 2021-02-05 | 北京邮电大学 | Mechanical arm trajectory tracking control method based on data driving |
| CN113635902A (en) * | 2021-08-23 | 2021-11-12 | 同济大学 | Active control method for pure electric start shaking of hybrid electric vehicle |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130268150A1 (en) * | 2012-04-04 | 2013-10-10 | Feisel Weslati | Predictive powertrain control using powertrain history and gps data |
| CN104401232A (en) * | 2014-12-21 | 2015-03-11 | 吉林大学 | Electric vehicle torque optimization method based on data driven predictive control |
| CN105221275A (en) * | 2015-08-20 | 2016-01-06 | 吉林大学 | Based on the Turbocharged Gasoline Engine control system of Model Distinguish |
| CN106080584A (en) * | 2016-06-21 | 2016-11-09 | 江苏大学 | A kind of hybrid vehicle pattern based on Model Predictive Control Algorithm switching control method for coordinating |
-
2017
- 2017-01-16 CN CN201710028468.5A patent/CN106647280B/en not_active Expired - Fee Related
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130268150A1 (en) * | 2012-04-04 | 2013-10-10 | Feisel Weslati | Predictive powertrain control using powertrain history and gps data |
| CN104401232A (en) * | 2014-12-21 | 2015-03-11 | 吉林大学 | Electric vehicle torque optimization method based on data driven predictive control |
| CN105221275A (en) * | 2015-08-20 | 2016-01-06 | 吉林大学 | Based on the Turbocharged Gasoline Engine control system of Model Distinguish |
| CN106080584A (en) * | 2016-06-21 | 2016-11-09 | 江苏大学 | A kind of hybrid vehicle pattern based on Model Predictive Control Algorithm switching control method for coordinating |
Non-Patent Citations (2)
| Title |
|---|
| 余开江 等: "混合动力汽车模型预测控制策略研究", 《***仿真技术》 * |
| 王剑 等: "基于模型预测控制的混合动力汽车转矩协调控制方法研究", 《新型工业化》 * |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107168319A (en) * | 2017-06-01 | 2017-09-15 | 同济大学 | A kind of unmanned vehicle barrier-avoiding method based on Model Predictive Control |
| CN108181811A (en) * | 2017-12-18 | 2018-06-19 | 长春工业大学 | A kind of slip rate tracking and controlling method based on linear time-varying |
| CN108181811B (en) * | 2017-12-18 | 2020-10-13 | 长春工业大学 | Slip rate tracking control method based on linear time variation |
| CN111301397A (en) * | 2020-02-28 | 2020-06-19 | 南京航空航天大学 | Variable time domain model prediction energy management method for plug-in hybrid electric vehicle |
| CN111301397B (en) * | 2020-02-28 | 2023-09-26 | 南京航空航天大学 | Method for managing prediction energy of variable time domain model of plug-in hybrid electric vehicle |
| CN112318511A (en) * | 2020-11-02 | 2021-02-05 | 北京邮电大学 | Mechanical arm trajectory tracking control method based on data driving |
| CN113635902A (en) * | 2021-08-23 | 2021-11-12 | 同济大学 | Active control method for pure electric start shaking of hybrid electric vehicle |
| CN113635902B (en) * | 2021-08-23 | 2024-02-27 | 同济大学 | Active control method for pure electric starting shake of hybrid electric vehicle |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106647280B (en) | 2019-05-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| He et al. | An improved MPC-based energy management strategy for hybrid vehicles using V2V and V2I communications | |
| CN106647280A (en) | HEV (hybrid electric vehicle) mode switching data driving prediction controller design method and HEV mode switching data driving prediction controller design system | |
| Yang et al. | A stochastic predictive energy management strategy for plug-in hybrid electric vehicles based on fast rolling optimization | |
| Yuan et al. | Intelligent energy management strategy based on hierarchical approximate global optimization for plug-in fuel cell hybrid electric vehicles | |
| Chen et al. | Energy management strategy for power-split plug-in hybrid electric vehicle based on MPC and double Q-learning | |
| CN105365592B (en) | Method and system for estimating remaining endurance mileage of electric automobile and electric automobile | |
| CN107168104B (en) | Observer-based longitudinal speed control method for pure electric intelligent automobile | |
| Borhan et al. | Nonlinear model predictive control for power-split hybrid electric vehicles | |
| Xu et al. | Q-learning-based supervisory control adaptability investigation for hybrid electric vehicles | |
| CN104401232B (en) | Electric automobile torque optimization method based on data-driven PREDICTIVE CONTROL | |
| CN103112450B (en) | Real-time optimized control method for plug-in parallel hybrid electric vehicle | |
| Qi et al. | Generalization ability of hybrid electric vehicle energy management strategy based on reinforcement learning method | |
| Ngo | Gear shift strategies for automotive transmissions | |
| Li et al. | A speedy reinforcement learning-based energy management strategy for fuel cell hybrid vehicles considering fuel cell system lifetime | |
| Ouddah et al. | From offline to adaptive online energy management strategy of hybrid vehicle using Pontryagin’s minimum principle | |
| Zhang et al. | Comparative study of energy management in parallel hybrid electric vehicles considering battery ageing | |
| CN112561169A (en) | Hybrid MPC energy management method for parallel hybrid electric vehicle | |
| Liu et al. | Hierarchical speed planning and energy management for autonomous plug-in hybrid electric vehicle in vehicle-following environment | |
| CN112721907A (en) | Efficient energy management method of network-connected hybrid electric vehicle considering road gradient | |
| CN104925063A (en) | Model predictive control method for electromechanical composite transmission vehicle | |
| Shabbir | Control strategies for series hybrid electric vehicles | |
| Goodarzi et al. | Stability enhancement and fuel economy of the 4-wheel-drive hybrid electric vehicles by optimal tyre force distribution | |
| Schaut et al. | Predictive thermal management for an electric vehicle powertrain | |
| Han et al. | Hierarchical optimization of speed and gearshift control for battery electric vehicles using preview information | |
| Kong et al. | A data-driven energy management method for parallel PHEVs based on action dependent heuristic dynamic programming (ADHDP) model |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20190524 Termination date: 20200116 |