CN108417031A - A kind of intelligent parking berth reservation policy optimization method based on Agent emulation - Google Patents
A kind of intelligent parking berth reservation policy optimization method based on Agent emulation Download PDFInfo
- Publication number
- CN108417031A CN108417031A CN201810212500.XA CN201810212500A CN108417031A CN 108417031 A CN108417031 A CN 108417031A CN 201810212500 A CN201810212500 A CN 201810212500A CN 108417031 A CN108417031 A CN 108417031A
- Authority
- CN
- China
- Prior art keywords
- parking
- parking lot
- time
- vehicle
- reservation
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
- G08G1/0125—Traffic data processing
- G08G1/0129—Traffic data processing for creating historical data or processing based on historical data
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
- G08G1/0137—Measuring and analyzing of parameters relative to traffic conditions for specific applications
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/14—Traffic control systems for road vehicles indicating individual free spaces in parking areas
- G08G1/145—Traffic control systems for road vehicles indicating individual free spaces in parking areas where the indication depends on the parking areas
- G08G1/148—Management of a network of parking areas
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Traffic Control Systems (AREA)
- Navigation (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a kind of intelligent parking berths based on Agent emulation to preengage policy optimization method.The basic thought of the present invention is vehicle Agent by handling the traffic information received, updates dynamic attribute in conjunction with itself static attribute and preset logic rules, and the behaviors such as movement and the decision in road network are realized according to dynamic attribute.The present invention passes through simulation optimization means, the intelligent parking berth reservation strategy of urban central zone is optimized, optimize the ratio for preengaging berth in each parking lot in rear region by reasonable disposition, adjust user's parking demand, alleviate urban central zone parking resource to be unevenly distributed, the phenomenon that local parking lot " finding no parking space ", local parking lot " nobody shows any interest in ", and then Optimizing Urban Transportation.
Description
Technical field
The present invention relates to the parking positions emulated based on Agent to preengage policy optimization method, belongs to intelligent parking technology neck
Domain.
Background technology
Agentbased system theory and emulation technology are that current progress complication system emulation is most active, most have an impact
One of method, basic thought be by simulating real world, complication system being divided into corresponding Agent, with from
The upward mode in bottom starts with from the individual microscopic behavior of research, and then obtains system macroscopic behavior.
Traffic problems are substantially the processes for participating in each of traffic individual (vehicle) in road online mobile and decision.So
And traditional static traffic distribution and parking preference pattern interacts between can not reflecting vehicle, is the parking choosing of an entirety
Select probability.
Rational berth reservation strategy, which can play, to be improved berth occupancy, vitalizes slack resources, optimization peripheral path fortune
Row and other effects;The evils such as unreasonable berth reservation strategy can cause the parking stall wasting of resources instead, intensified contradiction between supply and demand
Fruit.Therefore, it is necessary to which reservation tactful effect in berth is analyzed and evaluated and is optimized.
Invention content
The purpose of the present invention is to provide a kind of intelligent parking berths based on Agent emulation to preengage policy optimization method.
The basic thought of the present invention is vehicle Agent by handling the traffic information received, in conjunction with itself static attribute and
Preset logic rules update dynamic attribute, and the behaviors such as movement and the decision in road network are realized according to dynamic attribute.
The present invention includes the following steps:
C1, Math condition design, specifically:
C11 initialization emulation road networks
Gather including start node set O, terminal node set D, parking lot node set P, intersection node C and section
S.Initialize connection attribute between all kinds of nodes, distance;Initialize section grade, traffic capacity Cs, free stream velocityIn advance
Load each section initial flow q of road networks(0), the travel speed v in each section of initial road network is calculated using BPR functionss(0), stroke
Time ts(0), it calculates as follows:
Wherein, αs、βsFor the BPR function parameters of section s, LsFor the length of section s.
C12 initializes parking lot
Include the capacity C in each parking lotp, the ratio η in berth can be preengagep, initial parking number xp(0), parking rate fp, stop
Vehicle sails out of rate dp, maximum endure queuing vehicle length mp, queue queuep, queue length lp.The then available pool of reservation section
DigitWherein [] is rounding symbol, the available Berth number of non-reservation sectionAssuming that initial time
The practical parking position occupation rate of reservation section and non-reservation section be it is equal, i.e.,
Wherein,For the berth occupation rate of reservation section parking stall,For the berth occupation rate of non-reservation section parking stall.
C13 initializes simulated environment
Including dummy spacings, current iteration number, maximum iteration, the ratio of initialization emulation OD and subscriber
ηc。
C2, intelligent parking reservation system Agent Building of Simulation Model, specifically:
C21 user's decision process
C211 parking selections
For non-subscriber:When user reaches the intersection in road network, user can carry out parking selection, then plan most
Short time driving path.
User first screens away from destination, walking distance within the scope of 500m and not to cross parking lot sequence parking
.The preference pattern that stops considers journey time, walking distance, estimated cruising time, estimated queuing time, Parking Fee, parking
The calculating of field selection disutility indicates as follows:
Wherein,Indicate parking choosings of the non-reservation vehicle c from current intersection b to parking lot p in the m times iteration
Select effectiveness, tbp(m) it is the most short running time from current intersection b to parking lot p in the m times iteration, dpdFor from parking lot p
To the walking distance of destination d,To be expected cruising time,For in the m times iteration parking lot p it is estimated
Queuing time, ζ are random entry, a1、a2、a3、a4、a6、a7Respectively running time, walking distance, estimated cruising time, estimated etc.
Wait for the significant coefficient of time, Parking Fee and random entry, and each coefficient is respectively less than 0.
According to maximization of utility principle, non-subscriber selects the maximum parking lot of disutility as target parking lot.
After determining target parking lot, non-subscriber calculates shortest time traveling road according to the traffic of current road network
Diameter updates associated vehicle attribute.
For subscriber:Parking selection is carried out when entering road network.Parking alternative costs factor include journey time,
Walking distance, reservation expense, Parking Fee, parking lot select the calculating of disutility to indicate as follows:
Wherein,Indicate parking selections of the reservation vehicle c from current intersection b to parking lot p in the m times iteration
Effectiveness, tbp(m) it is the most short running time from current intersection b to parking lot p in the m times iteration, dpdFor from parking lot p to
The walking distance of destination d,Parking lot p moment, that is, m+t is reached to be estimatedbpThe reservation rate at moment, ζ are random
, a1、a2、a5、a6、a7Respectively running time, walking distance, reservation expense, Parking Fee and random entry significant coefficient, and
Each coefficient is respectively less than 0.
According to maximization of utility principle, user selects the maximum parking lot of disutility as target parking lot, updates user
Reservation expense
C212 Path selections
User travels according to road practical operation situation according to the shortest time path, and in each intersection, user understands foundation
Dijkstra's algorithm carries out path optimization's selection.
C22 association attributes update
C221 vehicle locations update
In simulation process, each iteration is required for the real time position of more new vehicle Agent, to realize vehicle Agent
Movement in default simulated environment, to further obtain simulation result.
In each emulation, the magnitude of traffic flow on every section can be measured, and each road can be calculated in conjunction with BPR functions
The travel speed v of sectionsWith journey time ts.If vehicle Agent is located on section, according to the travel speed in place section to
It moves ahead further, until reaching intersection;If vehicle Agent is located at intersection, according to current road network operation conditions
Decision goes out most short travel time path, advances to next intersection of planning path.
C222 road network states update
According to vehicle attribute present position AcAnd BcJudge the section where each car, updates in current the m times iteration
The volume of traffic q in all sectionss(m), update road network state includes each link travel speed vs(m) and journey time ts(m)。
C223 parking queue updates
When non-reservation vehicle reaches parking lot, if reaching parking lot without vacant berth but the length l of queuing vehiclepIt can
When receiving, vehicle Agent enters queue waiting.When parking lot has vehicle to sail out of, the queuing to parking lot
Sequence queuepUpdate, wherein queue queuepRecord the parking lot the vehicle Agent waited in line number.It is false
It is sailed out of with Poisson distribution if the vehicle in each parking lot is sailed out of.In each iteration, there is the parking lot for being lined up situation having vehicle
After sailing out of, the berth of computation-free is stopped according to the car number sequence of queue, successively until parking lot is without vacant pool
The vehicle Agent of position, completion of stopping completes simulation process, is rejected from simulated environment, while by the queue in the parking lot
Update.
C23 records road network and parking lot state, judges end condition
C231 simulation status preserves
Preserve each link counting of road network, each parking lot reservation section and the practical parking for taking reservation section of each iteration
Number.
C232 judges end condition
If current iteration number is no more than maximum iteration, continue the iteration of step c2.
If reaching end condition, the correlation time attribute of all vehicles of batch updating and travel cost etc.:
Journey time is that the time in one parking lot of moment to first time arrival is generated from vehicle, i.e.,
Wherein, ttcFor journey time,The moment is generated for vehicle,At the time of a parking lot being reached for first time.
From the time to the last one parking lot of arrival at the time of first parking lot of arrival, i.e., search time is
Wherein, tscFor search time,At the time of to reach first parking lot,To reach the last one parking
At the time of field.
From the time completed to parking at the time of reaching the last one parking lot, i.e., stand-by period is
Wherein, twcFor the stand-by period,At the time of to reach the last one parking lot,At the time of completion for parking.
Total travel time of vehicle c is the sum of journey time, search time, stand-by period and walking time, i.e.,
tc=ttc+tsc+twc+tkc
Wherein, tcFor total travel time, ttc、tsc、twc、tkcRespectively journey time, search time, stand-by period and step
The row time.
Total travel cost of vehicle c is the sum of reservation expense and Parking Fee, i.e.,
Wherein, fcFor total travel cost of vehicle c,To preengage expense,For Parking Fee.
Policing parameter optimizations of the c3 based on optimization algorithm, specifically:
C31 determines Optimal Parameters
Manager come influence area user's parking behavior, and then influences system by the adjusting to parking supply and parking demand
System operational effect.
C32 determines object function
Optimization aim is divided into social benefit optimization and the optimization of parking lot benefit according to service object.Social benefit index includes
Journey time, stand-by period, cruising time, mileage travelled;Parking lot performance indicator includes parking lot income, parking position utilization
Time.
Optimization object function is:
Wherein, α is weight coefficient, λ1, λ2For equalizing coefficient,For user's average travel total time-consuming,Averagely stop for user
Fare is used, and z is optimization aim.
C33 carries out parameter optimization by optimization algorithm
Optimization algorithm is chosen after determining Optimal Parameters and optimization aim.
Beneficial effects of the present invention:The present invention is by simulation optimization means, to the intelligent parking berth of urban central zone
Reservation strategy optimizes, and optimizes the ratio for preengaging berth in each parking lot in rear region by reasonable disposition, adjusts user
Parking demand is alleviated urban central zone parking resource and is unevenly distributed, local parking lot " finding no parking space ", local parking lot " nothing
People makes inquiries " the phenomenon that, and then Optimizing Urban Transportation.
Description of the drawings
Fig. 1 is that parking position preengages policy optimization method flow chart;
Fig. 2 is Agent simulation contact surfaces;
Fig. 3 is the trip decision-making flow chart of subscriber;
Fig. 4 is the trip decision-making flow chart of non-subscriber;
Fig. 5 is case study on implementation road network schematic diagram;
Fig. 6 is using AVKT as the fitness curve graph of optimization aim;
Fig. 7 is the change curve for optimizing preceding parking lot berth occupation rate about the time;
Fig. 8 is change curve of the parking lot berth occupation rate about the time after optimization.
Specific implementation mode
Below in conjunction with attached drawing, the invention will be further described.
As depicted in figs. 1 and 2, basic step of the invention is as follows:
C1, Math condition design;
C2, intelligent parking reservation system Agent Building of Simulation Model;
C3, the policing parameter optimization based on optimization algorithm;
The detailed process of step c1 includes:
C11 initialization emulation road networks
Gather including start node set O, terminal node set D, parking lot node set P, intersection node C and section
S.Initialize connection attribute between all kinds of nodes, distance;Initialize section grade (trunk roads/secondary distributor road), traffic capacity Cs、
Free stream velocityPreload each section initial flow q of road networks(0), the stroke in each section of initial road network is calculated using BPR functions
Speed vs(0), journey time ts(0), it calculates as follows:
Wherein, αs、βsFor the BPR function parameters of section s, LsFor the length of section s.
C12 initializes parking lot
Include the capacity C in each parking lotp, the ratio η in berth can be preengagep, initial parking number xp(0), parking rate fp, stop
Vehicle sails out of rate dp, maximum endure queuing vehicle length mp, queue queuep, queue length lp.The then available pool of reservation section
DigitWherein [] is rounding symbol, the available Berth number of non-reservation sectionAssuming that initial time
The practical parking position occupation rate of reservation section and non-reservation section be it is equal, i.e.,
Wherein,For the berth occupation rate of reservation section parking stall,For the berth occupation rate of non-reservation section parking stall.
C13 initializes simulated environment
Including dummy spacings, current iteration number, maximum iteration, the ratio of initialization emulation OD and subscriber
ηc。
The detailed process of step c2 includes:
User on road network advances along given route in each iteration, and carries out decision, after all users execute, is
System update association attributes, and judge whether to terminate iteration.
C21 user's decision process
C211 parking selections
For non-subscriber:When user reaches the intersection in road network, user can carry out parking selection, then plan most
Short time driving path, is shown in Fig. 4.
User first screen it is away from destination walking distance within the scope of 500m and not to cross parking lot sequence parking
.The preference pattern that stops consider journey time, walking distance, estimated cruising time, estimated queuing time, Parking Fee etc. because
Element, parking lot select the calculating of disutility to indicate as follows:
Wherein,Indicate parking choosings of the non-reservation vehicle c from current intersection b to parking lot p in the m times iteration
Select effectiveness, tbp(m) it is most short running time in the m times iteration from current intersection b to parking lot p (by dijkstra's algorithm
Solution obtains), dpdFor from parking lot p to the walking distance of destination d,For for estimated cruising time,For
The estimated queuing time of parking lot p in the m times iteration, ζ are random entry, a1、a2、a3、a4、a6、a7Respectively running time, step
Row distance, the significant coefficient for being expected cruising time, estimated stand-by period, Parking Fee and random entry, and each coefficient is respectively less than 0.
The calculating for estimating cruising time is as follows:
WhereinFor the cruising time under freestream conditions, xp(m+tbp) it is to be stopped according to the estimated arrival that historical data is predicted
The parking lot p moment, that is, m+tbpThe parking number at moment, αp、βpRespectively cruising time model coefficient.
It is expected that queuing timeIt can be calculated according to the thought of queueing theory as follows:
Wherein apFor parking lot p largest cumulative stop number,It is averaged the storage period for the vehicle of parking lot p.
According to maximization of utility principle, non-subscriber selects the maximum parking lot of disutility as target parking lot.
After determining target parking lot, non-subscriber calculates shortest time traveling road according to the traffic of current road network
Diameter updates associated vehicle attribute.
For subscriber:Parking selection is carried out when entering road network, sees Fig. 3.The factor for alternative costs of stopping includes row
Journey time, walking distance, reservation expense, Parking Fee etc., parking lot selects the calculating of disutility to indicate as follows:
Wherein,Indicate parking selections of the reservation vehicle c from current intersection b to parking lot p in the m times iteration
Effectiveness, tbp(m) it (is asked by dijkstra's algorithm for the most short running time from current intersection b to parking lot p in the m times iteration
Solution obtains), dpdFor from parking lot p to the walking distance of destination d,Parking lot p moment, that is, m+t is reached to be estimatedbp
The reservation rate at moment, ζ are random entry, a1、a2、a5、a6、a7Respectively running time, walking distance, reservation expense, parking fee
With the significant coefficient with random entry, and each coefficient is respectively less than 0.
Reservation expenseWith estimate cruising timeIt is proportionate, then, preengage expensekpFor cost coefficient.
According to maximization of utility principle, user selects the maximum parking lot of disutility as target parking lot, updates user
Reservation expense
C212 Path selections
User can travel according to road practical operation situation according to the shortest time path, and in each intersection, user can be according to
Path optimization's selection is carried out according to dijkstra's algorithm.
C22 association attributes update
C221 vehicle locations update
In simulation process, each iteration is required for the real time position of more new vehicle Agent, to realize vehicle Agent
Movement in default simulated environment, to further obtain simulation result.
In each emulation, the magnitude of traffic flow on every section can be measured, and each road can be calculated in conjunction with BPR functions
The travel speed v of sectionsWith journey time ts.If vehicle Agent is located on section, according to the travel speed in place section to
It moves ahead further, until reaching intersection;If vehicle Agent is located at intersection, according to current road network operation conditions
Decision goes out most short travel time path, advances to next intersection of planning path.
C222 road network states update
According to vehicle attribute present position AcAnd BcJudge the section where each car, updates in current the m times iteration
The volume of traffic q in all sectionss(m), update road network state includes each link travel speed vs(m) and journey time ts(m)。
There is traffic capacity C in formulas, free stream velocityThe length L of section ss。
C223 parking queue updates
When non-reservation vehicle reaches parking lot, if reaching parking lot without vacant berth but the length l of queuing vehiclepIt can
When receiving, vehicle Agent enters queue waiting.When parking lot has vehicle to sail out of, need to parking lot
Queue queuepUpdate, wherein queue queuepRecord the parking lot the vehicle Agent waited in line volume
Number.Assuming that the vehicle in each parking lot is sailed out of and is sailed out of with Poisson distribution.In each iteration, there is the parking lot for being lined up situation to exist
After having vehicle to sail out of, the berth of computation-free is stopped according to the car number sequence of queue, successively until parking lot is without sky
The vehicle Agent in remaining berth, completion of stopping completes simulation process, is rejected from simulated environment, while by the queuing in the parking lot
Sequence updates.
C23 records road network and parking lot state, judges end condition
C231 simulation status preserves
Preserve each link counting of road network, each parking lot reservation section and the practical parking for taking reservation section of each iteration
Number.
C232 judges end condition
If current iteration number is no more than maximum iteration, continue the iteration of step c2.
If reaching end condition, the correlation time attribute of all vehicles of batch updating and travel cost etc.:
Journey time is that the time in one parking lot of moment to first time arrival is generated from vehicle, i.e.,
Wherein, ttcFor journey time,The moment is generated for vehicle,At the time of a parking lot being reached for first time.
From the time to the last one parking lot of arrival at the time of first parking lot of arrival, i.e., search time is
Wherein, tsc is search time,At the time of to reach first parking lot,To reach the last one parking
At the time of field.
From the time completed to parking at the time of reaching the last one parking lot, i.e., stand-by period is
Wherein, twc is the stand-by period,At the time of to reach the last one parking lot,At the time of completion for parking.
Total travel time of vehicle c is the sum of journey time, search time, stand-by period and walking time, i.e.,
tc=ttc+tsc+twc+tkc
Wherein, tcFor total travel time, ttc、tsc、twc、tkcRespectively journey time, search time, stand-by period and step
The row time.
Total travel cost of vehicle c is the sum of reservation expense and Parking Fee, i.e.,
Wherein, fcFor total travel cost of vehicle c,To preengage expense,For Parking Fee.Wherein, non-subscriber
Reservation expense be 0.
Policing parameter optimizations of the c3 based on optimization algorithm
C31 determines Optimal Parameters
For parking management side, the role served as should be adjusted parking supply (as built parking lot) and influence parking
Demand (such as price mechanism, the setting of reservation berth).Manager can be stopped by the adjusting to these parameters come influence area user
Garage is, and then influences running effect.
C32 determines object function
Optimization aim can be divided into social benefit optimization and the optimization of parking lot benefit according to service object.Social benefit index
Including journey time, stand-by period, cruising time, mileage travelled etc.;Parking lot performance indicator includes parking lot income, parking pool
Position utilizes time etc..
The corresponding weight of each target can also be given to consider, such as:To trip total time-consuming and parking lot income two
Target is combined, and optimization object function is:
Wherein, α is weight coefficient, λ1, λ2For equalizing coefficient,For user's average travel total time-consuming,Averagely stop for user
Fare is used, and z is optimization aim.
C33 carries out parameter optimization by optimization algorithm
Optimization algorithm is chosen after determining Optimal Parameters and optimization aim, such as:The searchings such as genetic algorithm optimization solution.
To obtain effect of optimization, evaluation can be compared to system effect before and after optimization.
Evaluation content includes:
(1) emulation terminates the vehicle ratio of parking failure.
(2) it cruises vehicle proportion.
(3) wait in line vehicle proportion.
(4) the average cruising time of all vehicles.
(5) average latency of all vehicles.
(6) (AVKT, Average Vehicle Kilometers of Travel, one kind can for average VKT Vehicle Kilometers of Travel
The index of effective evaluation VMT Vehicle-Miles of Travel, significant for reducing fuel consumption, reduction trip distance).
(7) average travel total time-consuming (including journey time, cruising time, stand-by period, walking time).
(8) average Parking Fee (including subscription price and parking fee).
(9) parking lot berth always utilizes the time.
Embodiment:By taking road network shown in fig. 5 and parking lot layout's schematic diagram as an example, implement the real-time berth reservation in this parking lot
Strategy.
1, Math condition design
Simulation step length 1s is set, when emulation a length of 2h, i.e. 7200s.Total OD amounts are 2600, and berthing capacity is respectively 180,
300,160,310,190,210,230,240 and 280, total Berth number be 2100, the vehicle rate of sailing out of be desired for 60/
3600s, parking lot maximum endure 1/10 that queue length is the non-reservation section berthing capacity in the parking lot.Due to the present embodiment
The case where simulated environment is more than supply for demand, therefore cooling time is set for 200s, i.e., preceding 7000s has vehicle by set probability
It generates, rear 200s is generated without vehicle, is cruised berth for vehicle.
The parameter alpha of roadlock BPR functionss、βsIt is demarcated according to real road traffic flow, for trunk roads, αIt is main=1, βIt is main=5;It is right
In secondary distributor road, αIt is secondary=0.8, βIt is secondary=4.
Parking is cruised (search) time model parameter alphap、βpSituation takes α respectivelyp=2, βp=4.03.
2, intelligent parking reservation system Agent Building of Simulation Model
2.1 user's decision processes
(1) parking selection
For non-subscriber, when user reaches the intersection in road network, user can carry out parking selection, then plan most
Short time driving path.
User first screens away from destination, walking distance within the scope of 500m and not to cross parking lot sequence parking
.The preference pattern that stops consider journey time, walking distance, estimated cruising time, estimated queuing time, Parking Fee etc. because
Element, parking lot select the calculating of disutility to indicate as follows:
Wherein,Indicate parking choosings of the non-reservation vehicle c from current intersection b to parking lot p in the m times iteration
Select effectiveness, tbp(m) it is most short running time in the m times iteration from current intersection b to parking lot p (by dijkstra's algorithm
Solution obtains), dpdFor from parking lot p to the walking distance of destination d,To be expected cruising time,For
The estimated queuing time of parking lot p in the m times iteration, ζ are random entry, a1、a2、a3、a4、a6、a7Respectively running time, walking
Distance, the significant coefficient for being expected cruising time, estimated stand-by period, Parking Fee and random entry, and each coefficient is respectively less than 0.
The calculating for estimating cruising time is as follows:
WhereinFor the cruising time under freestream conditions, xp(m+tbp) it is to be stopped according to the estimated arrival that historical data is predicted
The parking lot p moment, that is, m+tbpThe parking number at moment, αp、βpRespectively cruising time model coefficient.
It is expected that queuing timeIt can be calculated according to the thought of queueing theory as follows:
Wherein apFor parking lot p largest cumulative stop number,It is averaged the storage period for the vehicle of parking lot p.
According to maximization of utility principle, non-subscriber selects the maximum parking lot of disutility as target parking lot.
After determining target parking lot, non-subscriber calculates shortest time traveling road according to the traffic of current road network
Diameter updates associated vehicle attribute.
For subscriber:The factor for alternative costs of stopping includes journey time, walking distance, reservation expense, parking fee
With etc., parking lot selects the calculating of disutility to indicate as follows:
Wherein,Indicate parking selections of the reservation vehicle c from current intersection b to parking lot p in the m times iteration
Effectiveness, tbp(m) it (is asked by dijkstra's algorithm for the most short running time from current intersection b to parking lot p in the m times iteration
Solution obtains), dpdFor from parking lot p to the walking distance of destination d,Parking lot p moment, that is, m+t is reached to be estimatedbp
The reservation rate at moment, ζ are random entry, a1、a2、a5、a6、a7Respectively running time, walking distance, reservation expense, parking fee
With the significant coefficient with random entry, and each coefficient is respectively less than 0.
Reservation expenseWith estimate cruising timeIt is proportionate, then, preengage expensekpFor cost coefficient.
According to maximization of utility principle, user selects the maximum parking lot of disutility as target parking lot, updates user
Reservation expense
(2) Path selection
User can travel according to road practical operation situation according to the shortest time path, and in each intersection, user can be according to
Path optimization's selection is carried out according to dijkstra's algorithm.
2.2 association attributes update
(1) vehicle location is updated
If vehicle is located on the last a bit of section on Ordinary Rd section or before parking lot, according to road grid traffic shape
Condition is along the sections of road certain distance, distance Fs of the more new vehicle c away from next intersection (or node)c。
Fc(m)=max { [Fc(m-1)-vs(m-1)], 0 }
Meanwhile the actual travel mileage K of more new vehicle cc=min { vs(m-1), Fc(m-1)}。
(2) road network state is updated
According to vehicle attribute present position AcAnd BcJudge the section where each car, updates in current the m times iteration
The volume of traffic q in all sectionss(m), update road network state includes each link travel speed vs(m) and journey time ts(m)。
There is traffic capacity C in formulas, free stream velocityThe length L of section ss。
(3) queue is updated
Judge whether parking lot has queuing, if lp>0, illustrate that parking lot p has vehicle queue waiting, judges that the parking lot is
It is no to have vacant berth, if there is vacant berth x, queue queuepPreceding y vehicle complete parking, picked from simulated environment
It removes, wherein y takes min (x, lp)。
2.3 record road networks and parking lot state, judge end condition
(1) simulation status preserves
Preserve each link counting of road network, each parking lot reservation section and the practical parking for taking reservation section of each iteration
Number.
(2) judge end condition
If current iteration number is no more than maximum iteration, continue to update next time.
If reaching end condition, batch updating institute vehicle attribute in need.
3, the policing parameter optimization based on optimization algorithm
3.1 determine Optimal Parameters
This example is in fixed subscriber ratio ηcUnder the premise of=0.5, survey region parking lot can preengage berth Set scale
γpOptimum organization problem, each parking lot just begin reservation berth ratio be all 0.5.
3.2 determine object function
Average VKT Vehicle Kilometers of Travel AVKT is an index for weighing vehicle cluster mileage travelled situation in certain region, in phase
With under conditions of, AVKT is smaller, and it is fewer to illustrate that vehicle arrives at required total kilometres, cruises, reduce for reduction
Regional traffic pressure is significant.
The present embodiment is with the minimum optimization aims of AVKT.
3.3 carry out parameter optimization by optimization algorithm
Since 9 parking lots can preengage berth ratio γpInfluence to system performance is not independent, influences each other
, since calculation amount is huge, can not searching optimization solution directly be compared by enumerative technique, this example is using common heuristic calculation
Method --- genetic algorithm optimizes object solving.
Population initial value is set as:[0.5,0.5,0.5,0.5,0.5,0.5,0.5,0.5,0.5], i.e., each parking lot can
It is initially 0.5 to preengage berth ratio;Crossing-over rate is 0.85, aberration rate 0.15, population quantity 20, iterative algebra 50
In generation, carries out simulation optimization to AVKT, as shown in Figure 6 as the fitness curve of optimization aim using AVKT, it is seen that optimization aim AVKT
Optimized with genetic algebra increase and is gradually tended to be steady.
Initial AVKT is 1695m, and the fitness value of simulation optimization is 1569m, reduces 7.4%.Each parking lot is preengage
Berth ratio is respectively:0.4,0.3,0.2,0.8,0.2,0.5,0.1,0.5,0.8, the capacity in comprehensive each parking lot, in region
The 48% of total berthing capacity in the berth occupied area domain that can be preengage.
4, optimization is front and back relatively
The front and back parking lot berth rate change curve of optimization is as shown in Figure 7 and Figure 8, it is seen then that before optimization, " popular parking lot "
3,5,7 after emulation starts berth occupation rate rise rapidly, and " unexpected winner parking lot " utilization rate deficiency if 9.
And the parking lot berth rate change curve after optimizing then shows that " popular parking lot " occupation rate rate of climb slows down,
And " unexpected winner parking lot " utilization rate obviously increases.
Comparison optimizes preceding parking lot berth rate change curve and preengages berth ratio with after optimization, it is known that:
Under AVKT this optimization aim, berth ratio very little is preengage in " popular parking lot " 3,5,7, and " unexpected winner
Parking lot " 4,9 to preengage berth ratio very high.This just illustrates, under the target for focusing on AVKT, needs to reduce popular parking lot
Preengage berth ratio, guide more reservation vehicles using other neighbouring parking lots, to reach AVKT optimizations.
Specific to simulation numerical, the average traffic mileage number of all vehicles drops to 1583m from 1697m, has dropped
6.69%, overall average trip, which takes from 653s, drops to 620s;Decline 5.13%;Average Parking Fee drops to from 8.78 yuan
8.24 yuan, parking lot income reduces 6.2%.
Claims (3)
1. it is a kind of based on Agent emulation intelligent parking berth preengage policy optimization method, it is characterised in that this method include with
Lower step:
C1, Math condition design, specifically:
C11 initialization emulation road networks
Including start node set O, terminal node set D, parking lot node set P, intersection node C and section set S;Just
Connection attribute, distance between all kinds of nodes of beginningization;Initialize section grade, traffic capacity Cs, free stream velocityIt preloads
Each section initial flow q of road networks(0), the travel speed v in each section of initial road network is calculated using BPR functionss(0), journey time
ts(0), it calculates as follows:
Wherein, αs、βsFor the BPR function parameters of section s, LsFor the length of section s;
C12 initializes parking lot
Include the capacity C in each parking lotp, the ratio η in berth can be preengagep, initial parking number xp(0), parking rate fp, parking sail out of
Rate dp, maximum endure queuing vehicle length mp, queue queuep, queue length lp;The then available Berth number of reservation sectionWherein [] is rounding symbol, the available Berth number of non-reservation sectionAssuming that initial time is preengage
Part and non-reservation section practical parking position occupation rate be it is equal, i.e.,
Wherein,For the berth occupation rate of reservation section parking stall,For the berth occupation rate of non-reservation section parking stall;
C13 initializes simulated environment
Including dummy spacings, current iteration number, maximum iteration, the ratio η of initialization emulation OD and subscriberc;
C2, intelligent parking reservation system Agent Building of Simulation Model, specifically:
C21 user's decision process
C211 parking selections
For non-subscriber:When user reaches the intersection in road network, user can carry out parking selection, then plan most in short-term
Between driving path;
User first screens away from destination, walking distance within the scope of 500m and not to cross parking lot sequence parking lot;
The preference pattern that stops considers journey time, walking distance, estimated cruising time, estimated queuing time, Parking Fee, parking lot choosing
The calculating for selecting disutility indicates as follows:
Wherein,Indicate parking selection effects of the non-reservation vehicle c from current intersection b to parking lot p in the m times iteration
With tbp(m) it is the most short running time from current intersection b to parking lot p in the m times iteration, dpdFor from parking lot p to mesh
Ground d walking distance,To be expected cruising time,For the estimated queuing of the parking lot p in the m times iteration
Time, ζ are random entry, a1、a2、a3、a4、a6、a7Respectively running time, walking distance, estimated cruising time, estimated when waiting for
Between, the significant coefficient of Parking Fee and random entry, and each coefficient is respectively less than 0;
According to maximization of utility principle, non-subscriber selects the maximum parking lot of disutility as target parking lot;Determine mesh
After marking parking lot, non-subscriber calculates shortest time driving path according to the traffic of current road network, updates associated vehicle
Attribute;
For subscriber:Parking selection is carried out when entering road network;The factor for alternative costs of stopping includes journey time, walking
Distance, reservation expense, Parking Fee, parking lot select the calculating of disutility to indicate as follows:
Wherein,Indicate that parkings of the reservation vehicle c from current intersection b to parking lot p in the m times iteration selects effectiveness,
tbp(m) it is the most short running time from current intersection b to parking lot p in the m times iteration, dpdFor from parking lot p to destination
The walking distance of d,Parking lot p moment, that is, m+t is reached to be estimatedbpThe reservation rate at moment, ζ are random entry, a1、a2、
a5、a6、a7Respectively running time, walking distance, reservation expense, Parking Fee and random entry significant coefficient, and each coefficient is equal
Less than 0;
According to maximization of utility principle, user selects the maximum parking lot of disutility as target parking lot, updates the pre- of user
About expense
C212 Path selections
User travels according to road practical operation situation according to the shortest time path, and in each intersection, user understands foundation
Dijkstra's algorithm carries out path optimization's selection;
C22 association attributes update
C221 vehicle locations update
In simulation process, each iteration is required for the real time position of more new vehicle Agent, to realize vehicle Agent pre-
If the movement in simulated environment, to further obtain simulation result;
In each emulation, the magnitude of traffic flow on every section can be measured, and each section can be calculated in conjunction with BPR functions
Travel speed vsWith journey time ts;If vehicle Agent is located on section, according to the travel speed in place section to move ahead
Further, until reaching intersection;If vehicle Agent is located at intersection, according to current road network operation conditions decision
Go out most short travel time path, advances to next intersection of planning path;
C222 road network states update
According to vehicle attribute present position AcAnd BcJudge the section where each car, updates in current the m times iteration and own
The volume of traffic q in sections(m), update road network state includes each link travel speed vs(m) and journey time ts(m);
C223 parking queue updates
When non-reservation vehicle reaches parking lot, if reaching parking lot without vacant berth but the length l of queuing vehiclepIt can connect
By when, vehicle Agent enter queue waiting;When parking lot has vehicle to sail out of, to the queue in parking lot
queuepUpdate, wherein queue queuepRecord the parking lot the vehicle Agent waited in line number;Assuming that each
The vehicle in a parking lot is sailed out of to be sailed out of with Poisson distribution;In each iteration, there is the parking lot for being lined up situation thering is vehicle to sail
From rear, the berth of computation-free, stop successively according to the car number sequence of queue, until parking lot is without vacant berth,
The vehicle Agent completed that stops completes simulation process, is rejected from simulated environment, while more by the queue in the parking lot
Newly;
C23 records road network and parking lot state, judges end condition
C231 simulation status preserves
Preserve each link counting of road network, each parking lot reservation section and the practical parking number for taking reservation section of each iteration;
C232 judges end condition
If current iteration number is no more than maximum iteration, continue the iteration of step c2;
If reaching end condition, the correlation time attribute of all vehicles of batch updating and travel cost etc.:
Journey time is that the time in one parking lot of moment to first time arrival is generated from vehicle, i.e.,
Wherein, ttcFor journey time,The moment is generated for vehicle,At the time of a parking lot being reached for first time;
From the time to the last one parking lot of arrival at the time of first parking lot of arrival, i.e., search time is
Wherein, tscFor search time,At the time of to reach first parking lot,To reach the last one parking lot
Moment;
From the time completed to parking at the time of reaching the last one parking lot, i.e., stand-by period is
Wherein, twcFor the stand-by period,At the time of to reach the last one parking lot,At the time of completion for parking;
Total travel time of vehicle c is the sum of journey time, search time, stand-by period and walking time, i.e.,
tc=ttc+tsc+twc+tkc
Wherein, tcFor total travel time, ttc、tsc、twc、tkcRespectively journey time, search time, stand-by period and walking when
Between;
Total travel cost of vehicle c is the sum of reservation expense and Parking Fee, i.e.,
Wherein, fcFor total travel cost of vehicle c,To preengage expense,For Parking Fee;
Policing parameter optimizations of the c3 based on optimization algorithm, specifically:
C31 determines Optimal Parameters
Manager come influence area user's parking behavior, and then influences system fortune by the adjusting to parking supply and parking demand
Row effect;
C32 determines object function
Optimization aim is divided into social benefit optimization and the optimization of parking lot benefit according to service object;Social benefit index includes stroke
Time, stand-by period, cruising time, mileage travelled;When parking lot performance indicator includes parking lot income, parking position utilization
Between;
Optimization object function is:
Wherein, α is weight coefficient, λ1, λ2For equalizing coefficient,For user's average travel total time-consuming,It is averaged parking fee for user
With z is optimization aim;
C33 carries out parameter optimization by optimization algorithm
Optimization algorithm is chosen after determining Optimal Parameters and optimization aim.
2. policy optimization method is preengage in a kind of intelligent parking berth based on Agent emulation according to claim 1, special
Sign is:
It is expected that queuing timeIt is calculated according to the thought of queueing theory as follows:
Wherein apFor parking lot p largest cumulative stop number,It is averaged the storage period for the vehicle of parking lot p.
3. policy optimization method is preengage in a kind of intelligent parking berth based on Agent emulation according to claim 1, special
Sign is:
Reservation expenseWith estimate cruising timeIt is proportionate, the calculating for estimating cruising time is as follows:
WhereinFor the cruising time under freestream conditions, xp(m+tbp) it is the estimated arrival parking lot p moment predicted according to historical data
That is m+tbpThe parking number at moment, αp、βpRespectively cruising time model coefficient, then, preengage expense
kpFor cost coefficient.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810212500.XA CN108417031B (en) | 2018-03-15 | 2018-03-15 | Intelligent parking berth reservation strategy optimization method based on Agent simulation |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810212500.XA CN108417031B (en) | 2018-03-15 | 2018-03-15 | Intelligent parking berth reservation strategy optimization method based on Agent simulation |
Publications (2)
Publication Number | Publication Date |
---|---|
CN108417031A true CN108417031A (en) | 2018-08-17 |
CN108417031B CN108417031B (en) | 2020-04-24 |
Family
ID=63131626
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810212500.XA Active CN108417031B (en) | 2018-03-15 | 2018-03-15 | Intelligent parking berth reservation strategy optimization method based on Agent simulation |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108417031B (en) |
Cited By (17)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109584603A (en) * | 2018-10-16 | 2019-04-05 | 贵州金汀科技股份有限公司 | A kind of parking stall intelligent recognition and bootstrap technique and system |
CN110335500A (en) * | 2019-08-01 | 2019-10-15 | 深圳市顶尖传诚科技有限公司 | The unmanned intelligent operation platform of shared parking based on position networking |
CN112863235A (en) * | 2021-01-05 | 2021-05-28 | 北京智联云海科技有限公司 | Method and device for predicting parking space vacancy rate |
CN113313327A (en) * | 2021-06-29 | 2021-08-27 | 浙江大学 | Agent simulation-based electric vehicle charging pile layout optimization method |
CN113434990A (en) * | 2021-06-29 | 2021-09-24 | 浙江大学 | Agent simulation-based electric car charging pile charging optimization method |
CN113515892A (en) * | 2021-06-24 | 2021-10-19 | 东南大学 | Multi-agent traffic simulation parallel computing method and device |
US20210350702A1 (en) * | 2019-02-07 | 2021-11-11 | Ningbo Geely Automobile Research & Development Co., Ltd. | System and method for guiding a vehicle occupant to an available vehicle parking |
CN113936498A (en) * | 2021-10-13 | 2022-01-14 | 上海辉略企业发展集团有限公司 | Parking resource dynamic configuration method based on multivariate data |
CN114117788A (en) * | 2021-11-25 | 2022-03-01 | 浙江大学 | Agent simulation method for parking transfer |
CN114298379A (en) * | 2021-12-06 | 2022-04-08 | 东南大学 | Automatic passenger-riding-substituting parking lot layout optimization method based on simulation optimization |
CN114519940A (en) * | 2022-02-25 | 2022-05-20 | 北京永利信达科技有限公司 | Big data analysis method and equipment applied to intelligent parking |
CN114627672A (en) * | 2020-12-09 | 2022-06-14 | 中国科学院深圳先进技术研究院 | Scenic spot tourist flow parking distribution method and system based on multiple intelligent agents |
CN114730525A (en) * | 2019-11-28 | 2022-07-08 | 邱海燕 | Parking space distribution system and method |
CN114758496A (en) * | 2022-03-30 | 2022-07-15 | 同济大学 | Method and device for guiding vehicles in parking lot and storage medium |
CN116167536A (en) * | 2022-12-07 | 2023-05-26 | 江苏巨楷科技发展有限公司 | Intelligent parking management method based on time period learning optimization |
CN117315926A (en) * | 2023-08-15 | 2023-12-29 | 广东省科学院广州地理研究所 | Parking prediction method and system based on traffic simulation model |
CN114627672B (en) * | 2020-12-09 | 2024-08-02 | 中国科学院深圳先进技术研究院 | Scenic spot tourist flow parking distribution method and system based on multiple intelligent agents |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR20060010956A (en) * | 2004-07-29 | 2006-02-03 | (주)한국공간정보통신 | Parking information system |
JP2006059056A (en) * | 2004-08-19 | 2006-03-02 | Nippon Telegr & Teleph Corp <Ntt> | Reservation method and system |
CN105489053A (en) * | 2015-11-19 | 2016-04-13 | 青岛智能产业技术研究院 | Establishment method for parallel parking system based on ACP method |
CN106355941A (en) * | 2016-09-27 | 2017-01-25 | 华中科技大学 | Intelligent parking lot system with parking space ordering function |
CN106652545A (en) * | 2015-10-28 | 2017-05-10 | 赵依军 | Method and system for optimizing processing of parking space reservation requests |
CN106846891A (en) * | 2017-03-02 | 2017-06-13 | 浙江大学 | A kind of Public Parking berth multistep forecasting method decomposed based on sequence |
-
2018
- 2018-03-15 CN CN201810212500.XA patent/CN108417031B/en active Active
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR20060010956A (en) * | 2004-07-29 | 2006-02-03 | (주)한국공간정보통신 | Parking information system |
JP2006059056A (en) * | 2004-08-19 | 2006-03-02 | Nippon Telegr & Teleph Corp <Ntt> | Reservation method and system |
CN106652545A (en) * | 2015-10-28 | 2017-05-10 | 赵依军 | Method and system for optimizing processing of parking space reservation requests |
CN105489053A (en) * | 2015-11-19 | 2016-04-13 | 青岛智能产业技术研究院 | Establishment method for parallel parking system based on ACP method |
CN106355941A (en) * | 2016-09-27 | 2017-01-25 | 华中科技大学 | Intelligent parking lot system with parking space ordering function |
CN106846891A (en) * | 2017-03-02 | 2017-06-13 | 浙江大学 | A kind of Public Parking berth multistep forecasting method decomposed based on sequence |
Non-Patent Citations (3)
Title |
---|
FUMITAKA KURAUCHI等: "Simulation Analysis on the Evaluation of Parking Reservation System", 《EASTS-JAPAN WORKING PAPER SERIES》 * |
LIYA GUO等: "A novel agent-based transportation model of a university campus with application to quantifying the environmental cost of parking search", 《TRANSPORTATION RESEARCH PART A》 * |
梅振宇 等: "基于停车选择行为的路内停车价格优化模型", 《交通运输***工程与信息》 * |
Cited By (23)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109584603A (en) * | 2018-10-16 | 2019-04-05 | 贵州金汀科技股份有限公司 | A kind of parking stall intelligent recognition and bootstrap technique and system |
US20210350702A1 (en) * | 2019-02-07 | 2021-11-11 | Ningbo Geely Automobile Research & Development Co., Ltd. | System and method for guiding a vehicle occupant to an available vehicle parking |
CN110335500A (en) * | 2019-08-01 | 2019-10-15 | 深圳市顶尖传诚科技有限公司 | The unmanned intelligent operation platform of shared parking based on position networking |
CN114730525A (en) * | 2019-11-28 | 2022-07-08 | 邱海燕 | Parking space distribution system and method |
CN114627672B (en) * | 2020-12-09 | 2024-08-02 | 中国科学院深圳先进技术研究院 | Scenic spot tourist flow parking distribution method and system based on multiple intelligent agents |
CN114627672A (en) * | 2020-12-09 | 2022-06-14 | 中国科学院深圳先进技术研究院 | Scenic spot tourist flow parking distribution method and system based on multiple intelligent agents |
CN112863235A (en) * | 2021-01-05 | 2021-05-28 | 北京智联云海科技有限公司 | Method and device for predicting parking space vacancy rate |
CN113515892B (en) * | 2021-06-24 | 2022-12-13 | 东南大学 | Multi-agent traffic simulation parallel computing method and device |
CN113515892A (en) * | 2021-06-24 | 2021-10-19 | 东南大学 | Multi-agent traffic simulation parallel computing method and device |
CN113313327A (en) * | 2021-06-29 | 2021-08-27 | 浙江大学 | Agent simulation-based electric vehicle charging pile layout optimization method |
CN113434990A (en) * | 2021-06-29 | 2021-09-24 | 浙江大学 | Agent simulation-based electric car charging pile charging optimization method |
CN113936498A (en) * | 2021-10-13 | 2022-01-14 | 上海辉略企业发展集团有限公司 | Parking resource dynamic configuration method based on multivariate data |
CN114117788A (en) * | 2021-11-25 | 2022-03-01 | 浙江大学 | Agent simulation method for parking transfer |
CN114298379A (en) * | 2021-12-06 | 2022-04-08 | 东南大学 | Automatic passenger-riding-substituting parking lot layout optimization method based on simulation optimization |
CN114298379B (en) * | 2021-12-06 | 2024-07-23 | 东南大学 | Automatic bus-substituting parking lot layout optimization method based on simulation optimization |
CN114519940A (en) * | 2022-02-25 | 2022-05-20 | 北京永利信达科技有限公司 | Big data analysis method and equipment applied to intelligent parking |
CN114519940B (en) * | 2022-02-25 | 2023-05-09 | 深圳市道网科技有限公司 | Big data analysis method and equipment applied to intelligent parking |
CN114758496B (en) * | 2022-03-30 | 2023-08-04 | 同济大学 | Induction method, device and storage medium for vehicles in parking lot |
CN114758496A (en) * | 2022-03-30 | 2022-07-15 | 同济大学 | Method and device for guiding vehicles in parking lot and storage medium |
CN116167536A (en) * | 2022-12-07 | 2023-05-26 | 江苏巨楷科技发展有限公司 | Intelligent parking management method based on time period learning optimization |
CN116167536B (en) * | 2022-12-07 | 2023-08-04 | 江苏巨楷科技发展有限公司 | Intelligent parking management method based on time period learning optimization |
CN117315926A (en) * | 2023-08-15 | 2023-12-29 | 广东省科学院广州地理研究所 | Parking prediction method and system based on traffic simulation model |
CN117315926B (en) * | 2023-08-15 | 2024-07-19 | 广东省科学院广州地理研究所 | Parking prediction method and system based on traffic simulation model |
Also Published As
Publication number | Publication date |
---|---|
CN108417031B (en) | 2020-04-24 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108417031A (en) | A kind of intelligent parking berth reservation policy optimization method based on Agent emulation | |
CN107766994B (en) | Shared bicycle scheduling method and scheduling system | |
CN107092976B (en) | Method for cooperatively optimizing departure intervals of multiple bus routes by multi-objective model | |
CN108764555A (en) | A kind of shared bicycle based on Hadoop parks a site selecting method | |
CN107640049A (en) | The mobile terminal system that a kind of electric automobile for charging station charges in order | |
CN108847037A (en) | A kind of city road network paths planning method towards non-global information | |
CN111667086B (en) | Vehicle ride-sharing path optimizing method and system | |
CN112330013B (en) | Electric vehicle charging guiding and pricing method based on dynamic road-electric coupling network | |
CN109612488B (en) | Big data micro-service-based mixed travel mode path planning system and method | |
CN112907103B (en) | Method for sharing dynamic supply and demand balance of single vehicle | |
CN114117700A (en) | Urban public transport network optimization research method based on complex network theory | |
CN107274035A (en) | A kind of transportation network and the method for electric automobile charging station coordinated planning | |
CN112418514A (en) | Method for optimizing campus bus route planning by using ant colony system | |
CN112418610A (en) | Charging optimization method based on fusion of SOC information and road network power grid information | |
CN111063191B (en) | Departure frequency and line network structure combined optimization method for bus line network design | |
CN115423168A (en) | Customized public transport network optimization method and system considering service stability | |
CN115510672A (en) | Electric micro-trip vehicle battery replacement demand prediction method under perception environment of Internet of things | |
Shi et al. | Deep q-network based route scheduling for transportation network company vehicles | |
CN116415882A (en) | Instant distribution order distribution system for collaborative distribution of riders and unmanned vehicles | |
CN112419716A (en) | Layout configuration method for shared single-vehicle facilities in track station transfer influence area | |
CN108830401B (en) | Dynamic congestion charging optimal rate calculation method based on cellular transmission model | |
CN111091286B (en) | Public bicycle scheduling method | |
CN112750298A (en) | Truck formation dynamic resource allocation method based on SMDP and DRL | |
CN112949987A (en) | Taxi dispatching and matching method, system, equipment and medium based on prediction | |
CN115907097A (en) | Hydrogenation station optimization layout method considering user travel characteristics and traffic flow |
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 |