CN107895960A - City rail traffic ground type super capacitor energy storage system energy management method based on intensified learning - Google Patents
City rail traffic ground type super capacitor energy storage system energy management method based on intensified learning Download PDFInfo
- Publication number
- CN107895960A CN107895960A CN201711053352.3A CN201711053352A CN107895960A CN 107895960 A CN107895960 A CN 107895960A CN 201711053352 A CN201711053352 A CN 201711053352A CN 107895960 A CN107895960 A CN 107895960A
- Authority
- CN
- China
- Prior art keywords
- energy
- storage system
- super capacitor
- train
- voltage
- 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
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J3/00—Circuit arrangements for ac mains or ac distribution networks
- H02J3/28—Arrangements for balancing of the load in a network by storage of energy
- H02J3/32—Arrangements for balancing of the load in a network by storage of energy using batteries with converting means
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J7/00—Circuit arrangements for charging or depolarising batteries or for supplying loads from batteries
- H02J7/34—Parallel operation in networks using both storage and other dc sources, e.g. providing buffering
- H02J7/345—Parallel operation in networks using both storage and other dc sources, e.g. providing buffering using capacitors as storage or buffering devices
Landscapes
- Engineering & Computer Science (AREA)
- Power Engineering (AREA)
- Charge And Discharge Circuits For Batteries Or The Like (AREA)
- Electric Propulsion And Braking For Vehicles (AREA)
Abstract
The present invention relates to a kind of city rail traffic ground type super capacitor energy storage system energy management method based on intensified learning.This method includes tactful netinit and on-line study two parts；Wherein tactful network initialization section utilizes known circuit, information of vehicles, the route map of train worked out in advance in city rail traffic, and the history vehicle data of actual acquisition, establishes more car Run-time scenario models；More car Run-time scenario models, floating voltage forecast model, direct current supply Load flow calculation algorithm and approximate dynamic programming algorithm are combined, obtain tactful network, the initial value as on-line study module；On-line study module uses model-free nitrification enhancement, and discharge and recharge threshold value on-line tuning is carried out by the method for super capacitor intelligent agent trial and error.The present invention can carry out on-line study in urban rail traction power supply net to super capacitor energy-storage system control strategy, realize the optimization of energy-saving effect and voltage regulation result.
Description
Technical field
The present invention relates to the control of track traffic and power-saving technology, specifically a kind of city rail traffic based on intensified learning
Face formula super capacitor energy-storage system capacity management method.
Background technology
In city rail traffic tractive power supply system, the pulse wave diode rectification of traction substation generally use 24 will
10kV/35kVAC alternating currents are converted into 750V/1500V direct currents, and haulage capacity is provided to circuit train.Due to diode rectification
With one-way, work as train braking, braking energy is delivered to Traction networks, if nearby being absorbed without tractor-trailer train, will make Traction networks
The rapid lifting of voltage, cause the generation of startup and the regeneration failure of braking resistor.In order to fully reclaim train regeneration energy, reduce
Regeneration failure and supply conductor voltage fluctuation, install super capacitor energy-storage system, as shown in figure 1, super capacitor in traction substation
Traction networks are connected to by two-way DC/DC, are in parallel with Rectification Power Factor.
Control strategy of the city rail traffic super capacitor energy-storage system generally use based on supply conductor voltage, including mode of operation
Selection and Double closed-loop of voltage and current two parts, respectively as shown in Fig. 2 (a), (b).The charging threshold of energy-storage system is set first
Value uch and discharge threshold uds, works as train traction, and electric substation's net forces down where energy-storage system puts in discharge threshold, super capacitor
Electricity, it is by Double closed-loop of voltage and current that supply conductor voltage is stable in discharge threshold；Work as train braking, supply conductor voltage is higher than
Charge threshold, super capacitor enter charge mode, reclaim the regenerating braking energy of train, and the stabilization for maintaining net to press.In order to protect
Card energy-storage system is stable, normally works, and sets super electricity in mode of operation selection and Double closed-loop of voltage and current respectively
Hold SOC and current limit, it is maintained within allowed band.
A kind of scheme of prior art is as shown in figure 3, to fix the control strategy of discharge and recharge threshold value.Energy storage system is set first
The charge threshold uch and discharge threshold uds of system, work as train traction, electric substation's net is forced down in discharge threshold where energy-storage system, is surpassed
Level electric capacity electric discharge, it is by Double closed-loop of voltage and current that supply conductor voltage is stable in discharge threshold；Work as train braking, Traction networks
Voltage is higher than charge threshold, and super capacitor enters charge mode, reclaims the regenerating braking energy of train, and it is steady to maintain net to press
It is fixed.The defects of this scheme is：Polytropy be present in the transportation condition of rail system：According to the passenger flow difference of daily different time sections,
Departure interval is adjusted according to service chart so that circuit train density changes；Bicycle power, dwell time exist certain
Randomness, therefore there is certain deviation in more car entirety operating modes and service chart；As the change of electricity need load, electric substation are empty
Carry voltage and slowly fluctuation occurs, influence the energy distribution between electric substation and energy-storage system.And prior art one is using fixation
The control strategy of threshold value, the change of transportation condition can not be adapted to, maintain good energy-conservation voltage regulation result；Can not be in energy-conservation voltage stabilizing effect
On-line tuning is carried out when fruit is not good enough, improves control effect.
Another scheme of prior art crosses journey as shown in figure 4, being based on specific double garages, energy-storage system is controlled excellent
Change problem regards a classical variational problem as, and isoperimetric constraint is considered in constraints, tries to achieve analytic solutions.The defects of this scheme
It is：Urban track traffic tractive power supply system is a complicated nonlinear time-varying network, it is difficult to its Accurate Model, solves storage
Bigger difficulty be present in the theoretical optimal control policy of energy system, and because model bias is difficult to obtain optimal control results.
In addition, the optimum results that prior art two obtains are only for specific running, it is impossible to suitable for circuit train operation work
Condition, departure interval continually changing actual scene.
The content of the invention
The technical problem to be solved by the invention is to provide a kind of city rail traffic ground type super electricity based on intensified learning
Hold energy-storage system energy management method, can in urban rail traction power supply net complexity time-varying, be difficult to Accurate Model in the case of, to super
Level capacitor energy storage system control strategy carries out on-line study, realizes the optimization of energy-saving effect and voltage regulation result；It is proposed based on reinforcing
The energy-storage system energy management method of study, the energy-conservation voltage regulation result optimization method as a kind of innovation.
The city rail traffic ground type super capacitor energy storage system energy management method based on intensified learning of the present invention, including
Tactful netinit and on-line study two parts；Wherein tactful network initialization section utilizes known line in city rail traffic
The history vehicle data of road, information of vehicles, the route map of train worked out in advance, and actual acquisition, establish more car Run-time scenarios
Model；More car Run-time scenario models, floating voltage forecast model, direct current supply Load flow calculation algorithm and approximate Dynamic Programming are calculated
Method combines, offline to solve energy-storage system optimal control problem, obtains tactful network, the initial value as on-line study module；Online
Study module uses model-free nitrification enhancement, and carrying out discharge and recharge threshold value by the method for super capacitor intelligent agent trial and error exists
Line adjusts, and the energy-conservation voltage regulation result of energy-storage system is optimized and is improved.
Further, more car Run-time scenario models, it is to use the overall operating mode of more car operations near energy-storage system
LSTM networks are predicted：It is primarily based on known circuit, vehicle parameter and route map of train and carries out traction calculating, obtains single-row
Velocity-time (V-t), power against time (P-t) and displacement versus time (S-t) sequence of car, in the difference of whole day route map of train
Period carries out sequential sampling (sequence length is the headway of place period), more car Run-time scenario sequences is obtained, such as formula
(1) shown in；
X (t)=[s1,p1,s2,p2,s3,p3,s4,p4] (1)；
Based on obtained sequence data initialization training LSTM networks；Then further according to the true train history of long-term record
Service data is adjusted to network parameter, makes its true train operating mode that more calculates to a nicety.
Further, the floating voltage forecast model, be by record electric substation's Rectification Power Factor electric current from 0 be changed on the occasion of
The output voltage at moment is period electric substation's floating voltage, whole day electric substation floating voltage change curve is obtained, with LSTM nets
Network is fitted.
Further, the tactful netinit is：
The Optimal Control Strategy of super capacitor energy-storage system, represent the form of an accepted way of doing sth (2)：
In formula, u (t) is decision variable, u (t)=[uch(t),uds(t)]；
J is control targe, considers the energy-conservation and voltage regulation result of energy-storage system herein, is defined as fractional energy savings e%
With net pressure improvement rate v% weighted sum, ω is weight coefficient.E% and v% calculation formula is respectively as shown in formula (3), (4)：
In formula (3),Represent to add respectively/plus during energy-storage system electric substation j voltage
And electric current；N represents electric substation's total quantity of statistics；Fractional energy savings e% is defined as electric substation before and after adding energy-storage system and always exports energy
Electric substation always exports the percentage of energy when amount variable quantity accounts for no energy-storage system；In formula (4),Respectively represent add/
The pantograph voltage of kth train when not adding energy-storage system；Nt is train sum in timing statisticses section and railroad section；Voltage stabilizing
Rate v% with train pantograph voltage beyond/assess less than the integration of certain value part.
Further, the on-line study module is：
Super capacitor EMS is considered as to the agency of study and decision-making, whole tractive power supply system, which is considered as, acts on behalf of institute
The environment at place；Agency obtains circuit train operation state, electric substation's state and itself SOC state by communicating, and performs corresponding
Action, so as to influence ambient condition and cause environment generation reflection energy-conservation, the prize signal of voltage regulation result；Agency obtains feedback
Action is improved after prize signal, by the optimization that sequential decision is realized with the mechanism of environmental interaction and trial and error；Including：
(a) state s, the displacement d of each train is includedk, power pk, wherein k expression kth trains, in addition to super capacitor
SOC states and electric substation's state, i.e. Rectification Power Factor electric current are changed into the output voltage u on the occasion of the moment from 0es；I.e.：
S=[d1,p1,…,dN,pN,soc,ues] (5)；
State set S is each train status set Straink, SOC state sets SSOCWith electric substation state SsubDirect product,
As shown in formula (6)；
S=Strain1×Strain2×…StrainN×Ssoc×Ssub(6)；
(b) a and tactful π is acted, energy-storage system action a is defined as the combination of discharge and recharge threshold value, i.e. a=[uds,uch]；Plan
Slightly π defines the behavior of agency, is mappings of the state set S to set of actions A：π：S→A；
(c) r is rewarded, prize signal is feedback of the environment to agent actions, and the target for acting on behalf of study obtains cumulative maximum
Reward；The reward of definition agency is time step Δ T internal segments energy rate, the increment of voltage improvement rate weighted sum, wherein weight coefficient
ω is taken as 0.5, as shown in formula (7)
R=-0.5 Δ v%-0.5 Δs e% (7)；
Progressive award meets relational expression (8) with energy-storage system control targe J；
J=1+r1+r2+…+rT (8)。
Advantages of the present invention is embodied in：
(1) nitrification enhancement is based on, super capacitor energy-storage system enters Mobile state adjustment to discharge and recharge threshold value, realizes different
Energy-conservation, the on-line optimization of voltage regulation result under the transportation conditions such as departure interval, floating voltage；
(2) tactful netinit is carried out based on city rail vehicle, circuit, service chart information and history data, so as to
Improve the learning efficiency of on-line learning algorithm；
Brief description of the drawings
Fig. 1 is the city rail traffic ground type super capacitor energy storage system schematic diagram of prior art；
Fig. 2 is the super capacitor control strategy schematic diagram based on voltage threshold of prior art；
Fig. 3 is the fixation discharge and recharge Hedging Point Control schematic diagram of prior art；
Fig. 4 is the energy-storage system parsing method for optimally controlling schematic diagram of prior art；
Fig. 5 is the energy-storage system energy management method schematic diagram based on intensified learning of the present invention；
Fig. 6 is the principle schematic of more car Run-time scenario models in the embodiment of the present invention；
Fig. 7 is the principle schematic of super capacitor intensified learning model in the embodiment of the present invention；
Fig. 8 is the pseudo-code figure of deterministic policy gradient method in the embodiment of the present invention.
Embodiment
This patent proposes the city rail traffic ground type super capacitor energy storage system energy management strategies based on intensified learning, by
Tactful network initialization module and on-line study module two parts composition, as shown in Figure 5.Wherein tactful network initialization section is filled
Divide and utilize known circuit, information of vehicles, the route map of train worked out in advance in city rail traffic, and the history car of actual acquisition
Data, establish more car Run-time scenario models；By more car Run-time scenario models, floating voltage forecast model, direct current supply trend
Computational algorithm and approximate dynamic programming algorithm combine, offline to solve energy-storage system optimal control problem, obtain tactful network, as
The initial value of on-line study module.Due to simulation model and the certain deviation of physical presence, and consider the change of actual motion condition,
On-line study module uses model-free nitrification enhancement, and discharge and recharge is carried out by the mechanism of super capacitor intelligent agent " trial and error "
Threshold value on-line tuning, the energy-conservation voltage regulation result of energy-storage system is set to be optimized and improved.
(1) more car Run-time scenario models
Fig. 6 is more car Run-time scenario models of this patent, by the overall operating mode LSTM of more car operations near energy-storage system
Network is predicted.It is primarily based on known circuit, vehicle parameter and route map of train and carries out traction calculating, obtains single vehicles
Velocity-time (V-t), power against time (P-t) and displacement versus time (S-t) sequence, in the different periods of whole day route map of train
Sequential sampling (sequence length is the headway of place period) is carried out, more car Run-time scenario sequences are obtained, such as formula (1) institute
Show.The train considered in the present embodiment is the train of the railroad section between adjacent two electric substation of energy-storage system.Based on obtaining
Sequence data initialization training LSTM networks；Then network is joined further according to the true train history data of long-term record
Number is adjusted, and makes its true train operating mode that more calculates to a nicety.
X (t)=[s1,p1,s2,p2,s3,p3,s4,p4] (1)
(2) floating voltage forecast model
Because electricity need load has greatly changed in whole day, there is fluctuation in electric substation's floating voltage, influence energy storage
The energy-conservation voltage regulation result of system.This patent is changed into the output voltage on the occasion of the moment by recording electric substation's Rectification Power Factor electric current from 0
For period electric substation's floating voltage, whole day electric substation floating voltage change curve is obtained, is fitted with LSTM networks.
(3) tactful netinit
The Optimal Control Strategy design of super capacitor energy-storage system is an Optimization for Sequential Decision Making problem, can represent an accepted way of doing sth
(2) form：
In formula, u (t) is decision variable, is herein energy-storage system discharge and recharge threshold value, i.e. u (t)=[uch(t),uds
(t)].The circuit equation that constraints contains tractive power supply system is constrained with the condition of work (electric current, SOC) of energy-storage system about
Beam.To ensure the stability and reliability of system operation, row constraint is entered by the mode of operation selection in Fig. 2 and double -loop control.
J is control targe, considers the energy-conservation and voltage regulation result of energy-storage system herein, is defined as fractional energy savings e%
With net pressure improvement rate v% weighted sum, ω is weight coefficient.E% and v% calculation formula is respectively as shown in formula (3), (4)：
In formula (3),Represent to add respectively/plus during energy-storage system electric substation j voltage
And electric current；N represents electric substation's total quantity of statistics.Therefore, it is always defeated to be defined as electric substation before and after adding energy-storage system by fractional energy savings e%
Go out the percentage that electric substation when energy variation amount accounts for no energy-storage system always exports energy.In formula (4),Represent respectively
Add/when not adding energy-storage system kth train pantograph voltage；Nt is that train is total in timing statisticses section and railroad section
Number.Therefore, voltage stabilizing rate v% herein with train pantograph voltage beyond/assess less than the integration of certain value part.
In tactful network initialization module, the control optimization problem passes through DC power flow analytical algorithm and approximate dynamic
Planning algorithm is implemented in combination with；And in on-line study module, using model-free nitrification enhancement, based on " attempt and failure " machine
System, study agency by with environmental interaction, obtain the feedback signal of evaluation property to obtain experience, carry out stragetic innovation, it is final real
The optimization of existing sequential decision.In order to accelerate on-line study speed, on-line study module is using the tactful network tried to achieve offline as initial value.
(4) on-line study module (super capacitor intensified learning model)
Fig. 7 is the theory diagram of super capacitor intensified learning model.By super capacitor EMS be considered as study and
The agency of decision-making, whole tractive power supply system are considered as the residing environment of agency.Agency obtains circuit train operation shape by communicating
State, electric substation's state and itself SOC state, corresponding action is performed, so as to influence ambient condition and cause environment generation reflection
Energy-conservation, the prize signal of voltage regulation result；Agency obtain feedback prize signal after to action be improved, by with environmental interaction
The optimization of sequential decision is realized with the mechanism of trial and error.
(a) state s.Including each train status (displacement dk, power pk, wherein k represents kth train), super capacitor shape
(Rectification Power Factor electric current is changed into the output voltage u on the occasion of the moment from 0 for state (super capacitor SOC) and electric substation's statees).I.e.：
S=[d1,p1,…,dN,pN,soc,ues] (5)
Therefore, state set S is each train status set Straink, SOC state sets SSOCWith electric substation state Ssub's
Direct product, as shown in formula (6).
S=Strain1×Strain2×…StrainN×Ssoc×Ssub (6)
(b) a and tactful π is acted.Energy-storage system action a is defined as the combination of discharge and recharge threshold value, i.e. a=[uds,uch]；Plan
Slightly π defines the behavior of agency, is mappings of the state set S to set of actions A：π：S→A.
(c) r is rewarded.Prize signal is feedback of the environment to agent actions, and the target for acting on behalf of study obtains cumulative maximum
Reward.The reward of definition agency herein is time step Δ T internal segments energy rate, the increment of voltage improvement rate weighted sum, wherein weight
Coefficient ω is taken as 0.5, as shown in formula (7)
R=-0.5 Δ v%-0.5 Δs e% (7)
Therefore, progressive award meets relational expression (8) with energy-storage system control targe J.Agency is according to super capacitor and row
Car state, the prize signal on voltage stabilizing, fractional energy savings is obtained, and discharge and recharge threshold value is improved on this basis, final
To optimal policy π*, make the maximized process of progressive award i.e. with solving the optimal control of energy-storage system by way of environmental interaction
Problem (2) processed, is optimal control targe J.
J=1+r1+r2+…+rT (8)
Super capacitor energy-storage system on-line learning algorithm uses depth deterministic policy gradient (DDPG) algorithm, is based on
Actor-critic (AC) learning framework, respectively with tactful network and value network approximation Strategy and value function, it is possible to achieve even
Continuous motion space control, and compared to randomized policy, it is necessary to which the data of sampling are few, efficiency of algorithm is high, therefore suitable for super
Capacitor charge and discharge threshold value continuously adjusts, and is advantageous to improve on-line study efficiency.
The pseudo-code of deterministic policy gradient method in order to solve deep neural network as shown in figure 8, when carrying out function approximation
The problem of nitrification enhancement is usually unstable, using experience replay and independent objective network.
Claims (5)
1. a kind of city rail traffic ground type super capacitor energy storage system energy management method based on intensified learning, it is characterized in that：
Including tactful netinit and on-line study two parts；Wherein tactful network initialization section utilizes known in city rail traffic
Circuit, information of vehicles, the route map of train worked out in advance, and the history vehicle data of actual acquisition, establish more car operation fields
Scape model；By more car Run-time scenario model, floating voltage forecast model, direct current supply Load flow calculation algorithm and approximate Dynamic Programmings
Algorithm combines, offline to solve energy-storage system optimal control problem, obtains tactful network, the initial value as on-line study module；
Line study module uses model-free nitrification enhancement, and discharge and recharge threshold value is carried out by the method for super capacitor intelligent agent trial and error
On-line tuning.
2. the city rail traffic ground type super capacitor energy storage system energy management according to claim 1 based on intensified learning
Method, it is characterized in that：More car Run-time scenario models, it is by the overall operating mode LSTM nets of more car operations near energy-storage system
Network is predicted：It is primarily based on known circuit, vehicle parameter and route map of train and carries out traction calculating, obtains the speed of single vehicles
Degree-time (V-t), power against time (P-t) and displacement versus time (S-t) sequence, enter in the different periods of whole day route map of train
Row sequential sampling, sequence length are the headway of place period, more car Run-time scenario sequences are obtained, as shown in formula (1)；
X (t)=[s1, p1, s2, p2, s3, p3, s4, p4] (1)；
Based on obtained sequence data initialization training LSTM networks；Then further according to the true train history run of long-term record
Data are adjusted to network parameter, make its true train operating mode that more calculates to a nicety.
3. the city rail traffic ground type super capacitor energy storage system energy management according to claim 1 based on intensified learning
Method, it is characterized in that：The floating voltage forecast model, be by record electric substation's Rectification Power Factor electric current from 0 be changed on the occasion of when
The output voltage at quarter is period electric substation's floating voltage, whole day electric substation floating voltage change curve is obtained, with LSTM networks
It is fitted.
4. further, the city rail traffic ground type super capacitance energy storage system according to claim 1 based on intensified learning
System energy management method, it is characterized in that：The tactful netinit is：
The Optimal Control Strategy of super capacitor energy-storage system, the form of an accepted way of doing sth (2) can be represented：
In formula, u (t) is decision variable, u (t)=[uch(t), uds(t)]；
J is control targe, considers the energy-conservation and voltage regulation result of energy-storage system herein, is defined as fractional energy savings e% and net
Improvement rate v% weighted sum is pressed, ω is weight coefficient.E% and v% calculation formula is respectively as shown in formula (3), (4)：
In formula (3),Represent to add respectively/plus during energy-storage system electric substation j voltage and electricity
Stream；N represents electric substation's total quantity of statistics；Fractional energy savings e% is defined as electric substation before and after adding energy-storage system and always exports energy quantitative change
Electric substation always exports the percentage of energy when change amount accounts for no energy-storage system；In formula (4),Represent to add/do not add respectively
The pantograph voltage of kth train when entering energy-storage system；Nt is train sum in timing statisticses section and railroad section；Voltage stabilizing rate
V% with train pantograph voltage beyond/assess less than the integration of certain value part.
5. the city rail traffic ground type super capacitor energy storage system energy management according to claim 1 based on intensified learning
Method, it is characterized in that：The on-line study module is：
Super capacitor EMS is considered as to the agency of study and decision-making, whole tractive power supply system is considered as residing for agency
Environment；Agency obtains circuit train operation state, electric substation's state and itself SOC state by communicating, and performs corresponding action,
So as to influence ambient condition and cause environment generation reflection energy-conservation, the prize signal of voltage regulation result；Agency obtains the reward of feedback
Action is improved after signal, by the optimization that sequential decision is realized with the mechanism of environmental interaction and trial and error；
Including：
(a) state s, the displacement d of each train is includedk, power pk, wherein k expression kth trains, in addition to super capacitor SOC
State and electric substation's state, i.e. Rectification Power Factor electric current are changed into the output voltage u on the occasion of the moment from 0es；I.e.：
S=[d1, p1..., dN, pN, soc, ues] (5)；
State set S is each train status set Straink, SOC state sets SSOCWith electric substation state SsubDirect product, such as formula
(6) shown in；
S=Strain1×Strain2×…StrainN×Ssoc×Ssub(6)；
(b) a and tactful π is acted, energy-storage system action a is defined as the combination of discharge and recharge threshold value, i.e. a=[uds, uch]；Tactful π determines
The behavior of justice agency, is mappings of the state set S to set of actions A：π：S→A；
(c) r is rewarded, prize signal is feedback of the environment to agent actions, and the target for acting on behalf of study obtains cumulative maximum reward；
The reward of definition agency is time step Δ T internal segments energy rate, the increment of voltage improvement rate weighted sum, and wherein weight coefficient ω is taken as
0.5, as shown in formula (7)
R=-0.5 Δ v%-0.5 Δs e% (7)；
Progressive award meets relational expression (8) with energy-storage system control targe J；
J=1+r1+r2+…+rT (8)。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711053352.3A CN107895960B (en) | 2017-11-01 | 2017-11-01 | Energy management method of urban rail transit energy storage system based on reinforcement learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711053352.3A CN107895960B (en) | 2017-11-01 | 2017-11-01 | Energy management method of urban rail transit energy storage system based on reinforcement learning |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107895960A true CN107895960A (en) | 2018-04-10 |
| CN107895960B CN107895960B (en) | 2020-03-17 |
Family
ID=61803880
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711053352.3A Active CN107895960B (en) | 2017-11-01 | 2017-11-01 | Energy management method of urban rail transit energy storage system based on reinforcement learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107895960B (en) |
Cited By (23)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109617207A (en) * | 2018-12-12 | 2019-04-12 | 北京交通大学 | Vehicle-mounted super capacitor tandem type topological structure and the control device of the structure, method |
| CN109672200A (en) * | 2018-12-11 | 2019-04-23 | 北京交通大学 | City rail traffic ground super capacitor energy-storage system distributed and coordinated control optimization method |
| CN110224387A (en) * | 2019-06-26 | 2019-09-10 | 西安许继电力电子技术有限公司 | A kind of urban rail super capacitor energy storage device charge and discharge control strategy |
| CN110460077A (en) * | 2019-08-27 | 2019-11-15 | 西南交通大学 | A kind of new energy system and control method for urban rail Traction networks |
| CN110889527A (en) * | 2018-09-07 | 2020-03-17 | 国网甘肃省电力公司电力科学研究院 | Electric vehicle charging load prediction method based on LSTM neural network |
| CN111237988A (en) * | 2020-01-15 | 2020-06-05 | 北京天泽智云科技有限公司 | Control method and system for subway vehicle-mounted air conditioning unit |
| CN111245034A (en) * | 2018-11-28 | 2020-06-05 | 兰州交通大学 | Capacity configuration scheme for ground type super-capacitor energy storage system of unmanned subway |
| CN111244982A (en) * | 2018-11-28 | 2020-06-05 | 兰州交通大学 | Capacity configuration scheme of rail transit ground type super capacitor energy storage system |
| CN111310384A (en) * | 2020-01-16 | 2020-06-19 | 香港中文大学（深圳） | Wind field cooperative control method, terminal and computer readable storage medium |
| CN111598434A (en) * | 2020-05-14 | 2020-08-28 | 重庆交通大学 | Method for determining configuration scheme of super-capacitor energy storage device of mountain urban rail transit |
| CN111628515A (en) * | 2020-06-18 | 2020-09-04 | 安徽能汇轨道交通技术有限公司 | Ground type hybrid energy storage system control method for rail transit |
| CN112606735A (en) * | 2020-12-22 | 2021-04-06 | 中车长春轨道客车股份有限公司 | Control method for optimal driving and energy management of non-contact power supply train |
| CN112731803A (en) * | 2021-04-01 | 2021-04-30 | 北京交通大学 | Energy storage system charging and discharging control method, device, equipment and readable storage medium |
| CN113054751A (en) * | 2021-03-25 | 2021-06-29 | 北京交通大学 | Coordination optimization method for urban rail transit vehicle-ground energy storage system based on information interaction |
| CN113619408A (en) * | 2021-08-30 | 2021-11-09 | 盾石磁能科技有限责任公司 | Power supply control method and device based on energy storage device, terminal and storage medium |
| CN113644671A (en) * | 2021-07-30 | 2021-11-12 | 湖南工业大学 | Dynamic power distribution control method for urban rail hybrid energy storage system based on deep reinforcement learning |
| CN113972714A (en) * | 2021-09-27 | 2022-01-25 | 湖南国天电子科技有限公司 | Energy-saving control method and system for heavy underwater robot super capacitor |
| CN115489320A (en) * | 2022-09-23 | 2022-12-20 | 西南交通大学 | Train pantograph intelligent control method based on deep reinforcement learning |
| CN115906622A (en) * | 2022-11-08 | 2023-04-04 | 杭州润氢科技有限公司 | Fuel cell electric vehicle energy control strategy based on model reinforcement learning |
| CN116169702A (en) * | 2023-04-26 | 2023-05-26 | 江苏为恒智能科技有限公司 | Large-scale energy storage prediction model and charging and discharging strategy based on reinforcement learning algorithm |
| CN116316755A (en) * | 2023-03-07 | 2023-06-23 | 西南交通大学 | Energy management method for electrified railway energy storage system based on reinforcement learning |
| CN116995641A (en) * | 2023-09-26 | 2023-11-03 | 北京交通大学 | Energy management architecture and method applied to rail transit energy storage |
| CN117955077A (en) * | 2024-03-27 | 2024-04-30 | 盾石磁能科技有限责任公司 | Charging and discharging control method, device and equipment of flywheel energy storage system |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106864276A (en) * | 2017-02-22 | 2017-06-20 | 薛建仁 | It is a kind of that to use super capacitor or super capacitor and energy-storage battery be the empty ferroelectricity origin system of electrical source of power |
| CN106877479A (en) * | 2015-12-11 | 2017-06-20 | 范洪国 | A kind of vehicle-mounted super capacitor energy-storage system control method |
-
2017
- 2017-11-01 CN CN201711053352.3A patent/CN107895960B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106877479A (en) * | 2015-12-11 | 2017-06-20 | 范洪国 | A kind of vehicle-mounted super capacitor energy-storage system control method |
| CN106864276A (en) * | 2017-02-22 | 2017-06-20 | 薛建仁 | It is a kind of that to use super capacitor or super capacitor and energy-storage battery be the empty ferroelectricity origin system of electrical source of power |
Non-Patent Citations (2)
| Title |
|---|
| 王彬等: "基于节能稳压的地面式超级电容储能系统容量配置优化研究", 《铁道学报》 * |
| 赵亚杰等: "基于动态阈值调节的城轨交通超级电容储能系统控制策略研究", 《电工技术学报》 * |
Cited By (34)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110889527A (en) * | 2018-09-07 | 2020-03-17 | 国网甘肃省电力公司电力科学研究院 | Electric vehicle charging load prediction method based on LSTM neural network |
| CN110889527B (en) * | 2018-09-07 | 2023-04-07 | 国网甘肃省电力公司电力科学研究院 | Electric vehicle charging load prediction method based on LSTM neural network |
| CN111244982A (en) * | 2018-11-28 | 2020-06-05 | 兰州交通大学 | Capacity configuration scheme of rail transit ground type super capacitor energy storage system |
| CN111245034A (en) * | 2018-11-28 | 2020-06-05 | 兰州交通大学 | Capacity configuration scheme for ground type super-capacitor energy storage system of unmanned subway |
| CN109672200B (en) * | 2018-12-11 | 2021-04-06 | 北京交通大学 | Distributed coordination control optimization method for urban rail transit ground super-capacitor energy storage system |
| CN109672200A (en) * | 2018-12-11 | 2019-04-23 | 北京交通大学 | City rail traffic ground super capacitor energy-storage system distributed and coordinated control optimization method |
| CN109617207A (en) * | 2018-12-12 | 2019-04-12 | 北京交通大学 | Vehicle-mounted super capacitor tandem type topological structure and the control device of the structure, method |
| CN109617207B (en) * | 2018-12-12 | 2021-04-06 | 北京交通大学 | Vehicle-mounted super capacitor series topology structure and control device and method thereof |
| CN110224387A (en) * | 2019-06-26 | 2019-09-10 | 西安许继电力电子技术有限公司 | A kind of urban rail super capacitor energy storage device charge and discharge control strategy |
| CN110460077A (en) * | 2019-08-27 | 2019-11-15 | 西南交通大学 | A kind of new energy system and control method for urban rail Traction networks |
| CN110460077B (en) * | 2019-08-27 | 2022-04-26 | 西南交通大学 | Control method of new energy power supply system for urban rail traction network |
| CN111237988A (en) * | 2020-01-15 | 2020-06-05 | 北京天泽智云科技有限公司 | Control method and system for subway vehicle-mounted air conditioning unit |
| CN111310384A (en) * | 2020-01-16 | 2020-06-19 | 香港中文大学（深圳） | Wind field cooperative control method, terminal and computer readable storage medium |
| CN111310384B (en) * | 2020-01-16 | 2024-05-21 | 香港中文大学（深圳） | Wind field cooperative control method, terminal and computer readable storage medium |
| CN111598434A (en) * | 2020-05-14 | 2020-08-28 | 重庆交通大学 | Method for determining configuration scheme of super-capacitor energy storage device of mountain urban rail transit |
| CN111598434B (en) * | 2020-05-14 | 2023-06-06 | 重庆交通大学 | Method for determining configuration scheme of super capacitor energy storage device of mountain urban rail transit |
| CN111628515A (en) * | 2020-06-18 | 2020-09-04 | 安徽能汇轨道交通技术有限公司 | Ground type hybrid energy storage system control method for rail transit |
| CN112606735B (en) * | 2020-12-22 | 2022-07-22 | 中车长春轨道客车股份有限公司 | Control method for optimal driving and energy management of non-contact power supply train |
| CN112606735A (en) * | 2020-12-22 | 2021-04-06 | 中车长春轨道客车股份有限公司 | Control method for optimal driving and energy management of non-contact power supply train |
| CN113054751A (en) * | 2021-03-25 | 2021-06-29 | 北京交通大学 | Coordination optimization method for urban rail transit vehicle-ground energy storage system based on information interaction |
| CN112731803A (en) * | 2021-04-01 | 2021-04-30 | 北京交通大学 | Energy storage system charging and discharging control method, device, equipment and readable storage medium |
| CN113644671A (en) * | 2021-07-30 | 2021-11-12 | 湖南工业大学 | Dynamic power distribution control method for urban rail hybrid energy storage system based on deep reinforcement learning |
| CN113619408A (en) * | 2021-08-30 | 2021-11-09 | 盾石磁能科技有限责任公司 | Power supply control method and device based on energy storage device, terminal and storage medium |
| CN113972714B (en) * | 2021-09-27 | 2023-08-08 | 湖南国天电子科技有限公司 | Energy-saving control method and system for super capacitor of heavy underwater robot |
| CN113972714A (en) * | 2021-09-27 | 2022-01-25 | 湖南国天电子科技有限公司 | Energy-saving control method and system for heavy underwater robot super capacitor |
| CN115489320A (en) * | 2022-09-23 | 2022-12-20 | 西南交通大学 | Train pantograph intelligent control method based on deep reinforcement learning |
| CN115906622A (en) * | 2022-11-08 | 2023-04-04 | 杭州润氢科技有限公司 | Fuel cell electric vehicle energy control strategy based on model reinforcement learning |
| CN116316755A (en) * | 2023-03-07 | 2023-06-23 | 西南交通大学 | Energy management method for electrified railway energy storage system based on reinforcement learning |
| CN116316755B (en) * | 2023-03-07 | 2023-11-14 | 西南交通大学 | Energy management method for electrified railway energy storage system based on reinforcement learning |
| CN116169702A (en) * | 2023-04-26 | 2023-05-26 | 江苏为恒智能科技有限公司 | Large-scale energy storage prediction model and charging and discharging strategy based on reinforcement learning algorithm |
| CN116169702B (en) * | 2023-04-26 | 2023-08-11 | 江苏为恒智能科技有限公司 | Large-scale energy storage charging and discharging method and model based on reinforcement learning algorithm |
| CN116995641A (en) * | 2023-09-26 | 2023-11-03 | 北京交通大学 | Energy management architecture and method applied to rail transit energy storage |
| CN116995641B (en) * | 2023-09-26 | 2023-12-22 | 北京交通大学 | Energy management architecture and method applied to rail transit energy storage |
| CN117955077A (en) * | 2024-03-27 | 2024-04-30 | 盾石磁能科技有限责任公司 | Charging and discharging control method, device and equipment of flywheel energy storage system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107895960B (en) | 2020-03-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107895960A (en) | City rail traffic ground type super capacitor energy storage system energy management method based on intensified learning | |
| Zou et al. | Reinforcement learning-based real-time energy management for a hybrid tracked vehicle | |
| Song et al. | A novel dual speed-curve optimization based approach for energy-saving operation of high-speed trains | |
| Liu et al. | A DQN-based intelligent control method for heavy haul trains on long steep downhill section | |
| Yan et al. | Hierarchical predictive energy management of fuel cell buses with launch control integrating traffic information | |
| Bai et al. | Energy-efficient locomotive operation for Chinese mainline railways by fuzzy predictive control | |
| Ning et al. | A synergistic energy-efficient planning approach for urban rail transit operations | |
| Wu et al. | Hierarchical predictive control for electric vehicles with hybrid energy storage system under vehicle-following scenarios | |
| Cipek et al. | Comparative analysis of conventional diesel-electric and hypothetical battery-electric heavy haul locomotive operation in terms of fuel savings and emissions reduction potentials | |
| CN107818383A (en) | A kind of optimization method and system of hybrid power train energy management strategies | |
| CN106056238B (en) | Planning method for train interval running track | |
| CN105667499A (en) | Energy management method for electric automobile in range extending mode | |
| CN109672200A (en) | City rail traffic ground super capacitor energy-storage system distributed and coordinated control optimization method | |
| CN109344996A (en) | A kind of urban railway transit train optimization and energy saving method | |
| CN103895533B (en) | For the method and relevant device of the operation for optimizing reversible traction substation | |
| CN109398426B (en) | Energy-saving driving strategy optimization method based on discrete ant colony algorithm under timing condition | |
| Yan et al. | Design of a deep inference framework for required power forecasting and predictive control on a hybrid electric mining truck | |
| Wang et al. | Real-time energy management strategy for a plug-in hybrid electric bus considering the battery degradation | |
| CN102897043A (en) | Method for allocating energy of extended-range type electric vehicle | |
| CN115534929A (en) | Plug-in hybrid electric vehicle energy management method based on multi-information fusion | |
| Wu et al. | Integrated energy management of hybrid power supply based on short-term speed prediction | |
| Pan et al. | Grey wolf fuzzy optimal energy management for electric vehicles based on driving condition prediction | |
| CN113644671A (en) | Dynamic power distribution control method for urban rail hybrid energy storage system based on deep reinforcement learning | |
| Zhang et al. | Energy management strategy optimization for hybrid energy storage system of tram based on competitive particle swarm algorithms | |
| Schenker et al. | Optimization model for operation of battery multiple units on partly electrified railway lines |
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 |