CN110435634A - A kind of stochastic dynamic programming energy management strategies optimization method based on diminution SOC feasible zone - Google Patents
A kind of stochastic dynamic programming energy management strategies optimization method based on diminution SOC feasible zone Download PDFInfo
- Publication number
- CN110435634A CN110435634A CN201910805940.0A CN201910805940A CN110435634A CN 110435634 A CN110435634 A CN 110435634A CN 201910805940 A CN201910805940 A CN 201910805940A CN 110435634 A CN110435634 A CN 110435634A
- Authority
- CN
- China
- Prior art keywords
- soc
- optimal trajectory
- feasible zone
- dynamic programming
- domain
- 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
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
- B60W20/00—Control systems specially adapted for hybrid vehicles
- B60W20/10—Controlling the power contribution of each of the prime movers to meet required power demand
- B60W20/15—Control strategies specially adapted for achieving a particular effect
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/80—Technologies aiming to reduce greenhouse gasses emissions common to all road transportation technologies
- Y02T10/84—Data processing systems or methods, management, administration
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Economics (AREA)
- Strategic Management (AREA)
- Human Resources & Organizations (AREA)
- Entrepreneurship & Innovation (AREA)
- Operations Research (AREA)
- Game Theory and Decision Science (AREA)
- Mechanical Engineering (AREA)
- Transportation (AREA)
- Automation & Control Theory (AREA)
- Marketing (AREA)
- Development Economics (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Electric Propulsion And Braking For Vehicles (AREA)
Abstract
The invention discloses a kind of based on the stochastic dynamic programming energy management strategies optimization method for reducing SOC feasible zone, includes the following steps: Step 1: obtaining the first SOC optimal trajectory under various working based on dynamic programming algorithm;And under finite time-domain, it is based on stochastic dynamic programming algorithm, obtains the 2nd SOC optimal trajectory under various working;Calculate separately the distance difference of the first SOC optimal trajectory and the corresponding state point of synchronization on the 2nd SOC optimal trajectory under various working;Step 2: being based on dynamic programming algorithm, the SOC optimal trajectory domain that global optimum's fuel economy is characterized under various working is obtained;Step 3: the width value in the SOC optimal trajectory domain under calculating separately various working, and according to width value, distance difference and the first SOC optimal trajectory, obtain SOC feasible zone.It is provided by the invention that required search condition point quantity in algorithm operational process can be reduced based on the stochastic dynamic programming energy management strategies optimization method for reducing SOC feasible zone, improve computational efficiency.
Description
Technical field
The invention belongs to vehicle energy management technical field, in particular to it is a kind of based on reduce SOC feasible zone with motor-driven
State plans energy management strategies optimization method.
Background technique
The Energy situation and environmental protection pressure increasingly serious in face of global range, development new-energy automobile, which has become society, to be held
The grand strategy and market new growth point of supervention exhibition.The one kind of oily electricity type hybrid vehicle as new energy vehicle, technology
Relative maturity, and market development has good prospects, and becomes best way of the traditional vehicle to new-energy automobile transition.As novel more
Powered vehicles, hybrid vehicle (hybrid electric vehicle, HEV) performance and its energy management control plan
It is slightly closely related.How according to predicting road conditions, power of vehicle demand, battery charge state (stage of charge, SOC) etc.
Information becomes to determine the operating mode that vehicle should be located to be reduced as far as equivalent fuel consumption and solves energy management control
The key of strategy.Currently, the energy management control strategy based on Dynamic Programming (dynamicprogramming, DP) algorithm can be with
The fuel economy of theoretically global optimum is obtained, as the standard for measuring other control strategy energy-saving effects.Since dynamic is advised
The realization for drawing control strategy is needed by known driving cycles information, and algorithm calculated load is big, cannot achieve real-time control,
Thus energy management control strategy of the exploitation based on stochastic dynamic programming algorithm.
Energy management method benefit based on stochastic dynamic programming (stochastic dynamic programming, SDP)
Existing standard condition or vehicle history running data is used to establish the statistical model of operating condition as the sample of stochastic model;Base
The energy management indicated with the statistical model is solved the problems, such as in dynamic programming algorithm.This method continues to use the Dynamic Programming multistage overall situation
Optimizing thought, introduces the statistical property of the following operating condition variation, and obtained optimal policy is that expected cost is most on a kind of average
Small strategy.Although Stochastic Dynamic Programming Method solves Dynamic Programming and needs asking for known vehicle overall situation work information in advance
Topic, but either finite time-domain SDP or infinite horizon SDP control strategy, practical application are often limited to the calculating of algorithm
Efficiency.Therefore, the calculated load of SDP algorithm how is effectively reduced to improve research heat of its computational efficiency as related fields
Point.The solution procedure of SDP control strategy can be construed to a kind of for generation multistage with discrete state amount, discrete control amount
Valence searching process, with the increase of the discrete precision of quantity of state, the calculating time of algorithm exponentially increases again.It can be seen that state
The quantity of variable is to influence the key factor of algorithm computational efficiency, thus, how to reduce required search shape in algorithm operational process
State quantity is the critical issue for needing to solve.
Summary of the invention
An object of the present invention is to provide a kind of stochastic dynamic programming based on diminution SOC feasible zone under finite time-domain
Energy management strategies optimization method reduces SOC feasible zone, according to the SOC optimal trajectory domain under different operating conditions to fast implement
The application on site of SDP control strategy under finite time-domain improves vehicle fuel economy.
The second object of the present invention is to provide a kind of stochastic dynamic programming based on diminution SOC feasible zone under infinite horizon
Energy management strategies optimization method is controlled on the basis of the SOC optimal trajectory that DP control strategy obtains according to DP under different operating conditions
The range difference of strategy and the SOC optimal trajectory of SDP control strategy reduces SOC feasible zone, to fast implement under infinite horizon
The application on site of SDP control strategy improves vehicle fuel economy.
Technical solution provided by the invention are as follows:
A kind of stochastic dynamic programming energy management strategies optimization method based on diminution SOC feasible zone, includes the following steps:
Step 1: being based on dynamic programming algorithm, the first SOC optimal trajectory and the various working under various working are respectively obtained
The SOC optimal trajectory domain of lower characterization global optimum's fuel economy;And under finite time-domain, calculated based on stochastic dynamic programming
Method obtains the 2nd SOC optimal trajectory under various working;
Step 2: calculating separately under various working on the first SOC optimal trajectory and the 2nd SOC optimal trajectory
The distance difference of the corresponding state point of synchronization;And calculate separately the width in the SOC optimal trajectory domain under various working
Angle value;
Step 3: it is feasible to obtain SOC according to the width value, the distance difference and the first SOC optimal trajectory
Domain.
Preferably, described based on the stochastic dynamic programming energy management strategies optimization method for reducing SOC feasible zone, also
Include:
Respectively to originate SOC and terminate SOC as critical point, pass through battery maximum charging current and battery maximum discharge current
It determines critical zone, obtains modified SOC feasible zone.
Preferably, the determination process of the critical zone includes:
Respectively to originate SOC, termination SOC as basic point, the straight line and battery of battery maximum charging current are characterized respectively
The straight line of maximum discharge current intersects with the SOC feasible zone up-and-down boundary, obtains modified SOC feasible zone;Wherein,
The straight slope of the characterization battery maximum charging current are as follows:
And the straight slope of characterization battery maximum charging current are as follows:
Wherein, Icharge, IdischargeRespectively battery maximum charging current and maximum discharge current, when battery charging,
Icharge< 0, when battery discharge, Idischarge> 0;QmaxFor battery capacity.
Preferably, in the step 2, the SOC for obtaining corresponding to global optimum's fuel economy under various working is optimal
The method in track domain, includes the following steps:
Step 1, the dynamic programming algorithm based on multistage optimizing obtain a plurality of 3rd SOC optimal trajectory, record all
The state point that three SOC optimal trajectories are passed through forms the set domain of optimal SOC point;
All SOC state points in original SOC feasible zone are numbered, and solve each moment pair in numerical order by step 2
All SOC discrete points answered and record and deposit the optimal trajectory where the discrete point at current time to the optimal value function of starting point
On previous moment SOC state point, until sequence operation to terminal;
Step 3, from terminal, the SOC state point on current time all optimal trajectories is successively searched and stored to backward,
Until backward is run to starting point, SOC optimal trajectory domain is formed.
Preferably, in the step 3, the method that obtains the SOC feasible zone are as follows:
The maximum width value L in the SOC optimal trajectory domain under multi-state is obtained respectively1、L2、……Ln, and calculate separately
S1、S2、……SnWith L1、L2、……LnDeviation ratio D1、D2、……Dn;Statistics obtains D1、D2、……DnIn maximum value Dmax,
And according to maximum value DmaxObtain amendment deviation ratio Damend;
Under i-th kind of operating condition, using the first SOC optimal trajectory as benchmark line, increase respectively in the two sides of the reference line
Add radius riWidth, obtain SOC feasible zone;
Wherein,ri=Li×(1+Damend), i=1,2 ... n;
And Damend>Dmax;
N indicates operating condition sum, S1、S2、……SnRespectively indicate the first SOC optimal trajectory and described second under various working
SOC optimal trajectory maximum distance difference.
Preferably, described based on the stochastic dynamic programming energy management strategies optimization method for reducing SOC feasible zone, also
It include:, using each discrete point at each moment as basic point, to be drawn using battery maximum discharge current as slope straight in the SOC feasible zone
Line obtains the discrete point of subsequent time.
Preferably, the various working includes: NEDC, UDDS, Japan 1015, FTP72 and HWEET operating condition.
Preferably, Damend=20%.
A kind of stochastic dynamic programming energy management strategies optimization method based on diminution SOC feasible zone, under infinite horizon,
Based on dynamic programming algorithm, the first SOC optimal trajectory under various working is obtained;And it is based on stochastic dynamic programming algorithm, it obtains
The 2nd SOC optimal trajectory under to various working;Calculate separately the first SOC optimal trajectory and described second under various working
The distance difference of the corresponding state point of synchronization on SOC optimal trajectory;And according to the distance difference and described first
SOC optimal trajectory obtains the SOC feasible zone of the stochastic dynamic programming algorithm under various working.
Preferably, the first SOC optimal trajectory and the 2nd SOC optimal trajectory under various working are obtained respectively
Maximum distance difference S1、S2、……Sn, count and obtain S1、S2、……SnIn maximum value Smax, and according to maximum value Smax
To the radius R of the region of search SOC;Under every kind of operating condition, using the first SOC optimal trajectory as benchmark line, in the reference line
Two sides increase separately the width of radius R, obtain the SOC feasible zone of stochastic dynamic programming algorithm;
Wherein, R > Smax。
The beneficial effects of the present invention are:
Stochastic dynamic programming energy management strategies optimization method provided by the invention based on diminution SOC feasible zone, is having
In limited time under domain, domain algorithm is sought based on the overall situation and obtains SOC optimal trajectory domain under DP control strategy, and forms ribbon on this basis
SOC feasible zone reduces finite time-domain SDP state search range, reduces required search condition point quantity in algorithm operational process, with
Improve the computational efficiency of algorithm;Being formed simultaneously based on SOC optimal trajectory domain obtained by DP control strategy can for the SDP SOC solved
Row domain ensure that the global suboptimality of hybrid vehicle energy management control strategy, improve the fuel-economy in global scope
Property;In addition, this method can obtain its optimal domain and the SOC feasible zone range for SDP offline according to different operating conditions, with quick
Realize the application on site of SDP control strategy.
Stochastic dynamic programming energy management strategies optimization method provided by the invention based on diminution SOC feasible zone, in nothing
In limited time under domain, on the basis of the SOC optimal trajectory that DP control strategy obtains, according to DP control strategy under different operating conditions and SDP away from
From SOC optimal trajectory range difference, reduce SOC feasible zone, the SDP control strategy under infinite horizon can be fast implemented
Application on site improves vehicle fuel economy.
Detailed description of the invention
Fig. 1 is the structure connection figure of stochastic dynamic programming energy management strategies optimization method of the present invention.
Fig. 2 is the schematic diagram of the track SOC distance computation obtained by DP and infinite horizon SDP under NEDC operating condition.
Fig. 3 is the schematic diagram of the track SOC distance computation obtained by DP and infinite horizon SDP under UDDS operating condition.
Fig. 4 is the schematic diagram in SDP control strategy state search domain under NEDC operating condition.
Fig. 5 is the schematic diagram in SDP control strategy state search domain under UDDS operating condition.
Fig. 6 is the schematic diagram of the track SOC distance computation obtained by DP and finite time-domain SDP under UDDS operating condition.
Fig. 7 is the schematic diagram in SDP control strategy state search domain under UDDS operating condition.
Fig. 8 is that the overall situation seeks domain algorithm routine explanatory diagram.
Fig. 9 is under NEDC operating condition based on the SOC optimal trajectory domain schematic diagram for improving DP formation.
Figure 10 is under UDDS operating condition based on the SOC optimal trajectory domain schematic diagram for improving DP formation.
Figure 11 is the schematic diagram that the optimal field width degree of SOC calculates under UDDS operating condition.
Figure 12 is the ribbon SOC feasible zone schematic diagram formed under UDDS operating condition based on SOC optimal trajectory domain.
Figure 13 is the schematic diagram of SOC feasible zone range optimization.
Figure 14 is the schematic diagram that artificial discrete SOC grid causes error.
Figure 15 is the discrete schematic diagram for improving SOC discrete domain.
Specific embodiment
Present invention will be described in further detail below with reference to the accompanying drawings, to enable those skilled in the art referring to specification text
Word can be implemented accordingly.
As shown in Figure 1, the present invention provides a kind of based on the stochastic dynamic programming energy management strategies for reducing SOC feasible zone
Optimization method, the specific implementation process is as follows:
One, the infinite horizon SDP optimization based on ribbon SOC feasible zone
Under infinite horizon, be based on Dynamic Programming (DP) and stochastic dynamic programming (SDP) algorithm, respectively obtain NEDC,
SOC optimal trajectory under the multiple groups operating condition such as UDDS, Japan 1015, FTP72, HWEET calculates SOC between two track of synchronization
The distance of state point.It counts under multiple groups operating condition based on distance between state point on SOC optimal trajectory obtained by DP and infinite horizon SDP
Maximum value, as shown in table 1.Wherein, by taking NEDC operating condition and UDDS operating condition as an example, as shown in Fig. 2, under NEDC operating condition based on DP with
The maximum distance computation schematic diagram of SOC optimal trajectory obtained by infinite horizon SDP;As shown in figure 3, under UDDS operating condition based on DP with
The maximum distance computation schematic diagram of SOC optimal trajectory obtained by infinite horizon SDP.
Maximum spacing statistical form based on SOC optimal trajectory obtained by DP and infinite horizon SDP under the different operating conditions of table 1
As shown in Table 1, under different operating conditions, shape on the SOC optimal trajectory that is obtained based on DP and infinite horizon SDP control strategy
Maximum distance is substantially in 0.12 range between state point.It follows that SOC optimal trajectory obtained by infinite horizon SDP control strategy is big
It causes to be located on the basis of DP control strategy SOC optimal trajectory, in the ribbon region formed with 0.12 for radius.If with the SOC
Feasible zone excludes extra SOC state point, SDP can be made in the search newly defined as SDP control strategy state search range
Optimizing is carried out in domain, and then reduces the calculated load of SDP algorithm, improves its computational efficiency.Meanwhile SOC required by DP control strategy
Optimal trajectory represents the corresponding global energy consumption optimal solution of this operating condition, provides accurate reference for SDP energy management strategies, guarantees SDP
The global suboptimality of energy management control strategy improves global fuel economy.
Therefore, it is based on the resulting SOC optimal trajectory of DP control strategy, using this track as core, for every on the track
One state point, width is the ribbon SDP control strategy state search domain formed in 0.12 range above and below the track, as nothing
The completely new state space that domain SDP is solved in limited time.By establishing this ribbon state search domain, SOC range is limited, is greatly contracted
The number of states of required search, helps quickly to obtain SDP control result, and then effectively change in small SDP algorithm solution procedure
Kind algorithm computational efficiency.By taking NEDC operating condition and UDDS operating condition as an example, as shown in figure 4, for SDP control strategy state under NEDC operating condition
Region of search;As shown in figure 5, for SDP control strategy state search domain under UDDS operating condition.
Two, based on the finite time-domain SDP optimization for improving SOC feasible zone
Step 1: obtaining ribbon SDP control strategy state search domain under finite time-domain
Referring to the forming process in ribbon SDP control strategy state search domain under infinite horizon, comparison finite time-domain SDP control
System strategy and the resulting SOC optimal trajectory of DP control strategy, calculate different operating conditions it is each when inscribe maximum distance between SOC point, choose
One upper limit value forms the state search domain for being directed to finite time-domain SDP control strategy.It counts DP under multiple groups difference operating condition and controls plan
Slightly and on the track SOC obtained by infinite horizon SDP control strategy between state point distance maximum value, as shown in table 2.Wherein, with
For UDDS operating condition, as shown in fig. 6, for the track SOC distance computation schematic diagram obtained by DP under UDDS operating condition and finite time-domain SDP.
The track SOC maximum spacing statistical form obtained by 2 finite time-domain SDP of table and DP
As shown in Table 2, under different operating conditions, the track SOC spacing obtained by finite time-domain SDP control strategy and DP control strategy is most
Big value basic guarantee is in 0.07 range.It as a result, can be by 0.07 width as finite time-domain SDP control strategy state search domain
Degree.It is similar with the forming process in ribbon infinite horizon SDP control strategy state search domain, most with SOC obtained by DP control strategy
Excellent track is core, forms ribbon SDP control strategy state search domain with 0.07 for upper and lower width.By taking UDDS operating condition as an example,
As shown in fig. 7, for SDP control strategy state search domain under UDDS operating condition.
Step 2: seeking domain algorithm based on the overall situation, SOC optimal trajectory domain under DP control strategy is formed
Based on conventional dynamic planning control strategy, vehicle fuel economy is optimal in available global sense one
The track SOC, corresponding optimal control sequence and consumption minimization value.In stage oil consumption calculating process, due to the calculating of stage oil consumption
It is related with the demand power of current generation, control amount and SOC variation, and SOC feasible zone discrete uniform, thus same stage exists
The SOC of many identical change amounts is shifted.Meanwhile the DP solution procedure based on multistage optimizing can be considered the tired of different phase cost
When accumulating and calculate, thus carrying out consumption minimization lookup based on global optimizing algorithm, finally obtained SOC optimal trajectory more than one.
Therefore, the exploitation overall situation seeks domain algorithm and obtains the track SOC of all characterization DP control strategy optimum control results, and algorithm routine is said
It is bright as shown in Figure 8.
All SOC optimal trajectories are obtained based on global optimizing algorithm and need the long period, and with the increase of output trajectory, are calculated
The method calculating time is multiplied, and therefore, it is difficult to realize the track SOC for exporting in the short time and all meeting the requirements.The present invention proposes one
The new optimal solution way of output of kind, i.e., " SOC optimal trajectory domain ".What this method was passed through by recording all SOC optimal trajectories
State point forms the set domain of optimal SOC point.The process for seeking " SOC optimal trajectory domain " mainly includes two parts: 1. will be owned
SOC state point renumbers, and sequentially solves the optimal value function of each SOC discrete point of each moment to starting point, and record the SOC from
All optimal trajectory institutes of scatterplot are through previous moment SOC state point, and corresponding storage is in a matrix, until sequence operation is to terminal;
2. from terminal, backward search and inscribed when storing each all optimal trajectory points (first look for terminal to it is previous when
The corresponding state point of the optimal trajectory at quarter, the corresponding state point of the optimal trajectory for searching the state point to its previous moment later,
And so on), until backward is run to starting point, the region that all optimum point SOC state points are formed is SOC optimal trajectory domain
(shadow region in Fig. 8).By taking NEDC operating condition and UDDS operating condition as an example, as shown in figure 9, for SOC optimal trajectory domain under NEDC operating condition;
It as shown in Figure 10, is SOC optimal trajectory domain under UDDS operating condition.
Traditional DP control is distributed in it is found that seeking based on the overall situation as Fig. 9 and Figure 10 required by the algorithm of domain " SOC optimal trajectory domain "
Below tactful gained SOC optimal trajectory (the resulting SOC optimal trajectory of DP control strategy i.e. in step 1), i.e. required by tradition DP
The optimal trajectory of SOC can be considered the upper contour line in " SOC optimal trajectory domain ".The reason is as follows that: during conventional dynamic programming evaluation,
Every SOC state point to terminal optimal energy consumption SOC track was sequentially calculated any time from top to bottom, when occurring in calculated result
When equal with the optimal power consumption values track SOC, the optimal track SOC does not update;And the overall situation seeks what hereafter algorithm record in domain occurred
The SOC state point that all optimal trajectories are passed through.
Step 3: being based on " SOC optimal trajectory domain ", ribbon SOC feasible zone is formed
Domain algorithm is sought by the overall situation, has obtained " the SOC optimal trajectory domain " of reflection global optimum's fuel economy.It calculates every
One moment region width shared by vertical direction, and using width maximum value shared by region as field width degree optimal under the operating condition,
To quantify the regional scope.Compare the width in gained ribbon SDP control strategy state search domain under finite time-domain and infinite horizon
Degree, it is known that, finite time-domain SDP control strategy state search domain range further reduces.By taking UDDS operating condition as an example, optimal field width degree
It calculates as shown in figure 11.
Seek the simulation run that domain algorithm completes multiple groups operating condition based on the overall situation, by its optimal field width degree and tradition DP and it is limited when
The SDP two kinds of track control strategy SOC maximum distance values in domain compare, and statistical result is as shown in table 3.
The optimal field width degree of SOC and the two kinds of strategy track SOC spacing contrast tables under the different operating conditions of table 3
As shown in Table 3, compared to the track SOC developed width required by traditional DP control strategy and SDP control strategy, SOC is most
There are a certain range of proportionate relationships for excellent domain width range and above-mentioned the two, and extent of deviation is in 20% range.By the optimal domain SOC
It is organically combined with statistical rules, ribbon SOC feasible zone is formed, as finite time-domain SDP energy management control strategy
Completely new state space.Specific practice are as follows: be primarily based on global domain algorithm of seeking and obtain SOC optimal trajectory domain and its width of corresponding operating condition
Angle value forms preliminary SOC feasible zone with optimal field width angle value on the basis of SOC optimal trajectory for upper and lower width, and in this range
On the basis of increase nearly 20%, to form the ribbon SOC feasible zone for solving SDP control strategy.By taking UDDS operating condition as an example, such as
Shown in Figure 12, for the ribbon SOC feasible zone formed based on SOC optimal trajectory domain.
Under different operating conditions, the SOC feasible zone tentatively established based on statistical method is compared and based on " the optimal domain SOC " side
The improvement SOC feasible zone range that method obtains, the results are shown in Table 4.
The SOC feasible zone range contrast table that two methods obtain under the different operating conditions of table 4
As shown in Table 4, by taking NEDC and UDDS operating condition as an example, statistical method is selected, the SOC for forming 0.07 width range is feasible
Domain;And selecting " the optimal domain DP " method formation width is respectively 0.054 and 0.020 SOC feasible zone, still meets SDP control
The distance range of strategy and the track SOC obtained by DP control strategy.Therefore, the SOC feasible zone obtained based on statistical rules is from big model
The SOC feasible zone for enclosing and optimizing state search range in angle, and obtained based on SOC optimal trajectory domain to the specific aim of operating condition more
By force, different range and more accurate SOC feasible zone can be obtained according to different operating conditions.
The memory space as needed for the solution of dynamic programming method is larger, and the calculating time is long, leads to algorithm computational efficiency
It is low, it is usually offline to complete, and gained control result is stored in controller.Therefore, in conjunction with mentioned above more careful, accurate
SOC feasible zone forming method can obtain offline its optimal domain and for SDP control strategy according to different operating conditions in practical applications
SOC feasible zone range, facilitate the application on site for fast implementing SDP control strategy.
Meanwhile the solution procedure of domain algorithm is sought it is found that the formation in SOC optimal trajectory domain will not be to DP control algolithm by the overall situation
The calculating time have an impact, i.e., to calculate equipment not will cause serious calculated load.Therefore, this method can be further excellent
Change finite time-domain SDP state search range, improves control algolithm computational efficiency and provide strong support.
Step 4: SOC feasible zone advanced optimizes, is formed and improve SOC feasible zone to optimize SDP control strategy
Formed ribbon SOC feasible zone after, for the ease of the application implementation of control algolithm, need to carry out state space from
It dissipates, obtains the SOC feasible zone of discretization.For finite time-domain SDP, the discrete way of state and discrete precision are to control strategy
Effect has vital effect.Therefore, it carries out carrying out finite time-domain SDP correlation SOC feasible zone in terms of following two excellent
Change:
(1) optimize SOC feasible zone range
The SOC feasible zone tentatively established is one using the track optimal SOC required by DP control strategy as the ribbon area of directrix
Domain.Since the limited battery capacity of hybrid vehicle needs to introduce battery permitted to extend the service life of battery
Maximum charging and discharging currents limitation.Based on this, respectively to originate SOC, termination SOC as basic point, draw respectively represent battery maximum fill/
The straight line of discharge current intersects with ribbon SOC feasible zone up-and-down boundary, forms new SOC feasible zone using as finite time-domain
The state search domain of SDP control strategy, as shown in figure 13.Wherein, the meter of the straight slope of battery maximum charging and discharging currents is characterized
It calculates as follows:
Wherein, ICharge,IdischargeRespectively battery maximum charge/discharge current (A), when battery charging, Icharge< 0,
When battery discharge, Idischarge> 0;QmaxFor battery capacity (Ah).
The new SOC feasible zone formed can further reduce SOC feasible zone range, reduce institute in SDP algorithm operational process
Search condition quantity is needed, to improve algorithm computational efficiency.
(2) optimize SOC feasible zone discrete way
The core of stochastic dynamic programming is a kind of multistage decision problem based on bellman principle, usually by state variable
It carries out uniformly discrete.It is attainable that this discrete way has ignored battery maximum electric discharge limitation (vehicle operation is in electric-only mode) institute
SOC point, thus have the defects that certain.As shown in figure 14, battery SOC shifts the position that can be reached under electric-only mode, with
There are error delta SOC for artificial discrete SOC grid, and leading to the calculating of stage oil consumption, there are errors, even result in finite field overall situation fuel oil
There are errors in computation for economy.Based on this, the present invention selects a kind of new SOC feasible zone discrete way, that is, requires to need that there are energy
Characterize vehicle operation in the SOC state point of pure electric vehicle drive mode, as shown in figure 15.Using each discrete point at each moment as basic point,
Using battery maximum discharge current as slope, the discrete point of subsequent time is obtained.When adjacent moment is 1s, δ SOCrealCalculating
It is as follows:
It obtains improving SOC feasible zone based on SOC feasible zone range and discrete optimization, as finite time-domain SDP energy management
The state search range of control strategy simplifies required search condition point quantity in traditional SDP control strategy solution procedure, to the greatest extent may be used
The computational efficiency of SDP algorithm can be improved to fast implement the application on site of SDP energy management control strategy.
Although the embodiments of the present invention have been disclosed as above, but its is not only in the description and the implementation listed
With it can be fully applied to various fields suitable for the present invention, for those skilled in the art, can be easily
Realize other modification, therefore without departing from the general concept defined in the claims and the equivalent scope, the present invention is simultaneously unlimited
In specific details and legend shown and described herein.
Claims (10)
1. a kind of based on the stochastic dynamic programming energy management strategies optimization method for reducing SOC feasible zone, which is characterized in that including
Following steps:
Step 1: being based on dynamic programming algorithm, the first SOC optimal trajectory and the various working following table under various working are respectively obtained
Levy the SOC optimal trajectory domain of global optimum's fuel economy;And under finite time-domain, it is based on stochastic dynamic programming algorithm, is obtained
The 2nd SOC optimal trajectory under to various working;
Step 2: calculating separately same on the first SOC optimal trajectory and the 2nd SOC optimal trajectory under various working
The distance difference of moment corresponding state point;And calculate separately the width value in the SOC optimal trajectory domain under various working;
Step 3: obtaining SOC feasible zone according to the width value, the distance difference and the first SOC optimal trajectory.
2. the stochastic dynamic programming energy management strategies optimization method according to claim 1 based on diminution SOC feasible zone,
It is characterized by further comprising:
Respectively to originate SOC and terminate SOC as critical point, determined by battery maximum charging current and battery maximum discharge current
Critical zone obtains modified SOC feasible zone.
3. the stochastic dynamic programming energy management strategies optimization method according to claim 2 based on diminution SOC feasible zone,
It is characterized in that, the determination process of the critical zone includes:
Respectively to originate SOC, termination SOC as basic point, straight line and the battery for being characterized battery maximum charging current respectively are maximum
The straight line of discharge current intersects with the SOC feasible zone up-and-down boundary, obtains modified SOC feasible zone;Wherein,
The straight slope of the characterization battery maximum charging current are as follows:
And the straight slope of characterization battery maximum charging current are as follows:
Wherein, Icharge, IdischargeRespectively battery maximum charging current and maximum discharge current, when battery charging, Icharge
< 0, when battery discharge, Idischarge> 0;QmaxFor battery capacity.
4. the stochastic dynamic programming energy management plan based on diminution SOC feasible zone according to claim 1 to 3
Slightly optimization method, which is characterized in that in the step 2, obtain corresponding to global optimum's fuel economy under various working
The method in SOC optimal trajectory domain, includes the following steps:
Step 1, the dynamic programming algorithm based on multistage optimizing obtain a plurality of 3rd SOC optimal trajectory, record all thirds
The state point that SOC optimal trajectory is passed through forms the set domain of optimal SOC point;
Step 2 numbers all SOC state points in original SOC feasible zone, and it is corresponding to solve each moment in numerical order
All SOC discrete points record and store on the optimal trajectory where the discrete point at current time to the optimal value function of starting point
Previous moment SOC state point, until sequence operation is to terminal;
Step 3, from terminal, the SOC state point on current time all optimal trajectories is successively searched and stored to backward, until
Backward is run to starting point, forms SOC optimal trajectory domain.
5. the stochastic dynamic programming energy management strategies optimization method according to claim 4 based on diminution SOC feasible zone,
It is characterized in that, in the step 3, the method that obtains the SOC feasible zone are as follows:
The maximum width value L in the SOC optimal trajectory domain under multi-state is obtained respectively1、L2、……Ln, and calculate separately S1、
S2、……SnWith L1、L2、……LnDeviation ratio D1、D2、……Dn;Statistics obtains D1、D2、……DnIn maximum value Dmax, and
And according to maximum value DmaxObtain amendment deviation ratio Damend;
Under i-th kind of operating condition, using the first SOC optimal trajectory as benchmark line, half is increased separately in the two sides of the reference line
Diameter riWidth, obtain SOC feasible zone;
Wherein,ri=Li×(1+Damend), i=1,2 ... n;
And Damend>Dmax;
N indicates operating condition sum, S1、S2、……SnRespectively indicate under various working the first SOC optimal trajectory and the 2nd SOC most
Excellent track maximum distance difference.
6. the stochastic dynamic programming energy management strategies optimization method according to claim 5 based on diminution SOC feasible zone,
It is characterized by further comprising:, using each discrete point at each moment as basic point, being discharged in the SOC feasible zone with battery maximum electric
Stream is that slope draws straight line, obtains the discrete point of subsequent time.
7. the stochastic dynamic programming energy management strategies optimization method according to claim 6 based on diminution SOC feasible zone,
It is characterized in that, the various working includes: NEDC, UDDS, Japan 1015, FTP72 and HWEET operating condition.
8. the stochastic dynamic programming energy management strategies optimization method according to claim 7 based on diminution SOC feasible zone,
It is characterized in that, Damend=20%.
9. a kind of based on the stochastic dynamic programming energy management strategies optimization method for reducing SOC feasible zone, which is characterized in that be based on
Dynamic programming algorithm obtains the first SOC optimal trajectory under various working;And under infinite horizon, advised based on stochastic and dynamic
Cost-effective method obtains the 2nd SOC optimal trajectory under various working;Calculate separately the first SOC optimal trajectory under various working
The distance difference of state point corresponding with the synchronization on the 2nd SOC optimal trajectory;And according to the distance difference
And the first SOC optimal trajectory, obtain the SOC feasible zone of the stochastic dynamic programming algorithm under various working.
10. according to claim 9 based on the stochastic dynamic programming energy management strategies optimization side for reducing SOC feasible zone
Method, which is characterized in that respectively obtain various working under the first SOC optimal trajectory and the 2nd SOC optimal trajectory most
Big distance difference S1、S2、……Sn, count and obtain S1、S2、……SnIn maximum value Smax, and according to maximum value SmaxIt obtains
The radius R of the region of search SOC;Under every kind of operating condition, using the first SOC optimal trajectory as benchmark line, in the reference line
Two sides increase separately the width of radius R, obtain the SOC feasible zone of stochastic dynamic programming algorithm;
Wherein, R > Smax。
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910805940.0A CN110435634B (en) | 2019-08-29 | 2019-08-29 | Random dynamic programming energy management strategy optimization method based on reduced SOC feasible domain |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910805940.0A CN110435634B (en) | 2019-08-29 | 2019-08-29 | Random dynamic programming energy management strategy optimization method based on reduced SOC feasible domain |
Publications (2)
Publication Number | Publication Date |
---|---|
CN110435634A true CN110435634A (en) | 2019-11-12 |
CN110435634B CN110435634B (en) | 2020-09-25 |
Family
ID=68438213
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910805940.0A Expired - Fee Related CN110435634B (en) | 2019-08-29 | 2019-08-29 | Random dynamic programming energy management strategy optimization method based on reduced SOC feasible domain |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110435634B (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111091249A (en) * | 2019-12-30 | 2020-05-01 | 吉林大学 | Method for realizing optimal distribution of vehicle global energy based on global domain-searching algorithm |
CN111740433A (en) * | 2020-06-26 | 2020-10-02 | 东北电力大学 | Energy storage primary frequency modulation control method based on dynamic droop coefficient and SOC recovery base point |
CN117470306A (en) * | 2023-10-31 | 2024-01-30 | 上海永大菌业有限公司 | Mushroom shed growth environment monitoring and analyzing method and system |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2001076027A (en) * | 1999-09-08 | 2001-03-23 | Nec Corp | Method and apparatus for simulation in system desgining stage and storage medium with program stored thereof |
CN102717797A (en) * | 2012-06-14 | 2012-10-10 | 北京理工大学 | Energy management method and system of hybrid vehicle |
CN102729987A (en) * | 2012-06-20 | 2012-10-17 | 浙江大学 | Hybrid bus energy management method |
CN108128302A (en) * | 2017-11-22 | 2018-06-08 | 吉林大学 | For the battery charge state planing method of hybrid vehicle global energy management |
WO2018160489A1 (en) * | 2017-03-01 | 2018-09-07 | Intel Corporation | Applying framing rules for a high speed data link |
US20190215729A1 (en) * | 2018-03-15 | 2019-07-11 | Intel Corporation | Session description protocol mechanisms for signaling radio access network capabilities in multimedia telephony sessions |
-
2019
- 2019-08-29 CN CN201910805940.0A patent/CN110435634B/en not_active Expired - Fee Related
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2001076027A (en) * | 1999-09-08 | 2001-03-23 | Nec Corp | Method and apparatus for simulation in system desgining stage and storage medium with program stored thereof |
CN102717797A (en) * | 2012-06-14 | 2012-10-10 | 北京理工大学 | Energy management method and system of hybrid vehicle |
CN102729987A (en) * | 2012-06-20 | 2012-10-17 | 浙江大学 | Hybrid bus energy management method |
WO2018160489A1 (en) * | 2017-03-01 | 2018-09-07 | Intel Corporation | Applying framing rules for a high speed data link |
CN108128302A (en) * | 2017-11-22 | 2018-06-08 | 吉林大学 | For the battery charge state planing method of hybrid vehicle global energy management |
US20190215729A1 (en) * | 2018-03-15 | 2019-07-11 | Intel Corporation | Session description protocol mechanisms for signaling radio access network capabilities in multimedia telephony sessions |
Non-Patent Citations (2)
Title |
---|
巴懋霖: "基于多工况优化的插电式混合动力汽车控制策略研究", 《中国优秀硕士学位论文全文数据库》 * |
邱利宏: "智能网联混合动力汽车队列分层优化控制方法研究", 《中国优秀硕士学位论文全文数据库》 * |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111091249A (en) * | 2019-12-30 | 2020-05-01 | 吉林大学 | Method for realizing optimal distribution of vehicle global energy based on global domain-searching algorithm |
CN111740433A (en) * | 2020-06-26 | 2020-10-02 | 东北电力大学 | Energy storage primary frequency modulation control method based on dynamic droop coefficient and SOC recovery base point |
CN117470306A (en) * | 2023-10-31 | 2024-01-30 | 上海永大菌业有限公司 | Mushroom shed growth environment monitoring and analyzing method and system |
CN117470306B (en) * | 2023-10-31 | 2024-04-02 | 上海永大菌业有限公司 | Mushroom shed growth environment monitoring and analyzing method and system |
Also Published As
Publication number | Publication date |
---|---|
CN110435634B (en) | 2020-09-25 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110435634A (en) | A kind of stochastic dynamic programming energy management strategies optimization method based on diminution SOC feasible zone | |
CN109272176B (en) | Calculation method is predicted to platform area line loss per unit using K-means clustering algorithm | |
CN108334991A (en) | A kind of electric automobile charging station method and system for planning | |
CN109606137A (en) | Merge the multi-source power drive system economy optimization method of cost factors of limit life | |
CN112918330B (en) | Method for calculating optimal working state control strategy of fuel cell vehicle | |
CN109510234B (en) | Hybrid energy storage capacity optimal configuration method and device for micro-grid energy storage power station | |
CN104913438B (en) | Ice-storage system controls optimization method and system | |
CN112140942B (en) | Self-adaptive equivalent consumption minimized energy management method for fuel cell vehicle | |
CN110175780A (en) | A kind of electric automobile charging station site selecting method, system and relevant apparatus | |
CN111064214A (en) | Power distribution network optimal scheduling method based on electric vehicle two-stage rolling strategy | |
CN113988384A (en) | Energy storage capacity optimal configuration method for improving reliability of power distribution network | |
CN113859018B (en) | Hierarchical charge-discharge optimization control method for large-scale electric automobile group | |
CN114212005A (en) | Energy management method and device for fuel cell system | |
Bashash et al. | Optimizing demand response of plug-in hybrid electric vehicles using quadratic programming | |
CN107492901A (en) | A kind of distributed energy storage system real-time optimization method and device | |
Guo et al. | Research on energy management strategy of heavy-duty fuel cell hybrid vehicles based on dueling-double-deep Q-network | |
CN115764849A (en) | Hybrid energy storage capacity optimal configuration method and configuration system thereof | |
CN109754123B (en) | Rotation center line distance weighting alternative positioning method for feeder line power supply range division | |
CN106451504A (en) | Control method and device for configuration cost of hybrid energy storage system | |
CN112455291B (en) | Instantaneous optimal energy management method for fuel cell automobile composite power supply | |
CN109635999A (en) | A kind of power station dispatching method looked for food based on population-bacterium and system | |
CN109447233B (en) | Electric vehicle charging and discharging scheduling method and system | |
CN111680469A (en) | Medium-low voltage distribution network line loss limit value calculation method | |
CN104408531B (en) | A kind of uniform dynamic programming method of multidimensional multistage complicated decision-making problems | |
CN110323779A (en) | A kind of method and system of the power dynamic aggregation of distributed power generation and energy storage device |
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: 20200925 Termination date: 20210829 |