CN1857981A - Group control lift dispatching method based on CMAC network - Google Patents
Group control lift dispatching method based on CMAC network Download PDFInfo
- Publication number
- CN1857981A CN1857981A CN200610040554.XA CN200610040554A CN1857981A CN 1857981 A CN1857981 A CN 1857981A CN 200610040554 A CN200610040554 A CN 200610040554A CN 1857981 A CN1857981 A CN 1857981A
- Authority
- CN
- China
- Prior art keywords
- beta
- elevator
- cmac
- parameter
- time
- 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
- VIEYMVWPECAOCY-UHFFFAOYSA-N 7-amino-4-(chloromethyl)chromen-2-one Chemical compound ClCC1=CC(=O)OC2=CC(N)=CC=C21 VIEYMVWPECAOCY-UHFFFAOYSA-N 0.000 title claims abstract description 20
- 238000000034 method Methods 0.000 title claims abstract description 10
- 230000003068 static effect Effects 0.000 claims abstract description 7
- 230000006870 function Effects 0.000 claims description 7
- 238000013016 damping Methods 0.000 claims description 6
- 238000004364 calculation method Methods 0.000 claims description 4
- 230000002787 reinforcement Effects 0.000 claims description 4
- 238000012546 transfer Methods 0.000 claims description 3
- 230000001105 regulatory effect Effects 0.000 abstract 1
- 238000002474 experimental method Methods 0.000 description 5
- 238000005516 engineering process Methods 0.000 description 4
- 238000005457 optimization Methods 0.000 description 3
- 238000011161 development Methods 0.000 description 2
- 230000003044 adaptive effect Effects 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000005265 energy consumption Methods 0.000 description 1
- 230000002068 genetic effect Effects 0.000 description 1
- 230000001537 neural effect Effects 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000000638 solvent extraction Methods 0.000 description 1
Images
Landscapes
- Elevator Control (AREA)
Abstract
The group controlled lift dispatching method based on CMAC network includes the following steps: 1. determining static parameters, dynamic parameters, passenger arrival model, CMAC network parameters and intensified learning parameters before triggering lift group control system; 2. observing lift state and calculating Q(x, run) and Q(x, stop); 3. selecting lift action; 4. setting the next decision point of each lift and renewing the lift R[i] value; 5. regulating the Q(s, a) estimations of each lift; 6. updating CMAC network weight value based on the expression; and 7. going to the step 1 after x-y and tx-ty. The present invention can raise lift dispatching efficiency and reduce the wait time of passengers.
Description
One, technical field
The present invention relates to a kind of dispatching method of elevator, relate in particular to a kind of elevator group controlling dispatching method.
Two, background technology
Individual calling ladder signal form is adopted in early stage elevator control, development along with computer controlled and intellectual technology, calling and command signal by one group of elevator of computing machine unified management, according to the optimization aim of default and the actual traffic situation in the building, produce optimum elevator dispatching strategy, Here it is present common multiple lift control system, the essence of its scheduling is to open, in the dynamic complex environment, a plurality of optimization aim such as passenger waiting time, passenger journey time, degree of congestion and energy consumption are being optimized control.At present the multiple control lift dispatching method relate generally to partitioning algorithm, based on the algorithm of search and rule-based algorithm or the like.Along with the development of intellectual technology, more and more researchers adopts the adaptive learning algorithms of technical study such as expert system, fuzzy control, artificial neural net (ANN) and genetic algorithm.But because elevator operates in the continuous time system, its state space higher-dimension, therefore external status perception and change and dynamically change with passenger's arrival rate is fully simultaneously effectively calculated optimal policy that elevator group controlling dispatches and is remained one of main difficult problem that research circle and industrial community face.
Consider that the actual environment that elevator faces is unknown, uncertain, and scheduling is the on-line optimization that arrives model at client.Therefore intensified learning (Reinforcement learning) technology is applied in the elevator group controlling scheduling, shows that by emulation experiment its method compares with at present existing calculation method, can obtain less client's average latency.
Three, summary of the invention
1, goal of the invention: the purpose of this invention is to provide a kind of efficient elevator group controlling dispatching method that can reduce passenger's average latency.
2, in order to reach above-mentioned goal of the invention, the present invention includes following step:
(1) determines that static parameter, dynamical parameter, passenger arrive model, CMAC network parameter and intensified learning parameter, trigger multiple lift control system then, wherein, static parameter is elevator number and floor number, dynamical parameter is that interfloor flight time, elevator stop/time to turn and passenger transfer time, the passenger arrives model parameter and is passenger's arrival-time distribution, the CMAC network parameter is input node, output node and extensive parameter, and the intensified learning parameter is exponential damping speed β and learning rate α;
(2) be located at t
xConstantly elevator i arrives a decision point, and observing and obtaining state is x, according to the CMAC network calculations draw Q (x, run) and Q (x, stop), wherein, Q (x, run) be the Q value function that elevator continues operation under the x state, (x stop) is the Q value function of elevator parking to Q;
(3) select action a according to following formula:
Wherein, T is temperature parameter and T>0;
(4) make the next decision point of elevator i occur in t
yConstantly, its corresponding state is y, according to formula
Upgrade the acquisition R[i of all elevators] value, wherein, R[i] be i portion elevator decision-making time point d[i from it] time begin total discount reinforcement value of accumulative total, t
0Be the time that a last incident takes place, t
1Be the time that current event takes place, for each at t
0And t
1Between actv. elevator-calling key b, make w
0(b) and w
1(b) be respectively t
0And t
1The time that passes after button b presses constantly, β is an exponential damping speed in the formula, and λ is client's a Poisson arrival rate;
(5) elevator i is according to formula:
(6) according to formula:
(7) with x ← y, t
x← t
y.. go to step 1, thereby realize the multiple control lift scheduling.
3, beneficial effect: its remarkable advantage is to reduce passenger's average latency effectively, improves the performance of elevator dispatching.
Table 1 only contains the contrast and experiment of down traffic pattern
Algorithm | AvgWait | SquaredWait | Percent>60s |
SECTOR | 21.4 | 674 | 1.12 |
RL-BP | 21.2 | 569 | 0.09 |
RL-CMAC | 19.7 | 529 | 0.07 |
Table 2 contains the contrast and experiment of uplink traffic pattern
Algorithm | AvgWait | SquaredWait | Percent>60s |
SECTOR | 27.3 | 1252 | 9.24 |
RL-BP | 24.3 | 1140 | 9.90 |
RL-CMAC | 21.8 | 1048 | 9.14 |
The contrast and experiment of table 3 twice uplink traffic
Algorithm | AvgWait | SquaredWait | Percent>60s |
SECTOR | 30.3 | 1643 | 13.50 |
RL-BP | 27.8 | 1698 | 8.74 |
RL-CMAC | 23.4 | 1562 | 8.20 |
Below under three kinds of transit modes, experimentize respectively, experiment shows based on the multiple control lift dispatching algorithm of CMAC network compares team control dispatching algorithm and the classical SECTOR algorithm of employing based on BP network intensified learning, can obtain less client's average latency; Client's wait time significantly reduces above the ratio of 60s simultaneously.
Four, description of drawings
Fig. 1 is the constructional drawing of intensified learning Function Estimation;
Fig. 2 is a CMAC neural network structure scheme drawing.
Five, the specific embodiment
As shown in Figure 1 and Figure 2, present embodiment comprises the following steps:
(1) determines that according to table 4 static parameter, dynamical parameter, passenger arrive model, CMAC network parameter and intensified learning parameter, trigger multiple lift control system then;
Table 4 embodiment parameter configuration
Parametric description | Parameter is provided with | |
Static parameter | The elevator number | 4 |
The floor number | 10 | |
Assignment key number in the ladder | 10 | |
Rated passenger capacity | 20 | |
Dynamical parameter | Interfloor flight time | 1.45s |
The elevator standing time | 7.19s | |
Elevator time to turn | 1s | |
Passenger transfer time | Average is that the erlang of 1s distributes, scope from 0.6 to 6.0s | |
Client arrives model | Time 00 | Arrival rate 1 |
Time 05 | Arrival rate 2 | |
Time 10 | Arrival rate 4 | |
Time 15 | Arrival rate 4 | |
Time 20 | Arrival rate 18 | |
Time 25 | Arrival rate 12 | |
Time 30 | Arrival rate 8 | |
Time 35 | Arrival rate 7 | |
Time 40 | Arrival rate 18 | |
Time 45 | Arrival rate 5 | |
Time 50 | Arrival rate 3 | |
Time 55 | Arrival rate 2 | |
The CMAC network parameter | The input node | 47 |
Output node | 2 | |
Extensive parameter | 3 | |
The intensified learning parameter | Exponential damping speed β | 0.01 |
Learning rate α | Satisfy ∑ α=∞, ∑ α 2<C<∞ |
(2) be located at t
xConstantly elevator i arrives a decision point, and observing and obtaining state is x, according to the CMAC network calculations draw Q (x, run) and Q (x, stop), wherein, Q (x, run) be the Q value function that elevator continues operation under the x state, (x stop) is the Q value function of elevator parking to Q;
(3) select action a according to following formula:
Wherein, T is temperature parameter and T>0;
(4) make the next decision point of elevator i occur in t
yConstantly, its corresponding state is y, according to formula
(5) elevator i is according to formula:
(6) according to formula:
(7) with x ← y, t
x← t
y.. go to step 1, thereby realize the multiple control lift scheduling.
Claims (1)
1, a kind of multiple control lift dispatching method based on the CMAC network is characterized in that this method may further comprise the steps:
(1) determines that static parameter, dynamical parameter, passenger arrive model, CMAC network parameter and intensified learning parameter, trigger multiple lift control system then, wherein, static parameter is elevator number and floor number, dynamical parameter is that interfloor flight time, elevator stop/time to turn and passenger transfer time, the passenger arrives model parameter and is passenger's arrival-time distribution, the CMAC network parameter is input node, output node and extensive parameter, and the intensified learning parameter is exponential damping speed β and learning rate α;
(2) be located at t
xConstantly elevator i arrives a decision point, and observing and obtaining state is x, according to the CMAC network calculations draw Q (x, run) and Q (x, stop), wherein, Q (x, run) be the Q value function that elevator continues operation under the x state, (x stop) is the Q value function of elevator parking to Q;
(3) select action a according to following formula:
Wherein, T is temperature parameter and T>0;
(4) make the next decision point of elevator i occur in t
yConstantly, its corresponding state is y, according to formula
Upgrade the acquisition R[i of all elevators] value, wherein, R[i] be i portion elevator decision-making time point d[i from it] time begin total discount reinforcement value of accumulative total, t
0Be the time that a last incident takes place, t
1Be the time that current event takes place, for each at t
0And t
1Between actv. elevator-calling key b, make w
0(b) and w
1(b) be respectively t
0And t
1The time that passes after button b presses constantly, β is an exponential damping speed in the formula, and λ is client's a Poisson arrival rate;
(5) elevator i is according to formula:
Regulate its Q (s, valuation a);
(6) according to formula:
(7) with x ← y, t
x← t
yGo to step 1, thereby realize the multiple control lift scheduling.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CNB200610040554XA CN100413771C (en) | 2006-05-24 | 2006-05-24 | Group control lift dispatching method based on CMAC network |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CNB200610040554XA CN100413771C (en) | 2006-05-24 | 2006-05-24 | Group control lift dispatching method based on CMAC network |
Publications (2)
Publication Number | Publication Date |
---|---|
CN1857981A true CN1857981A (en) | 2006-11-08 |
CN100413771C CN100413771C (en) | 2008-08-27 |
Family
ID=37296943
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CNB200610040554XA Expired - Fee Related CN100413771C (en) | 2006-05-24 | 2006-05-24 | Group control lift dispatching method based on CMAC network |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN100413771C (en) |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108408514A (en) * | 2018-03-14 | 2018-08-17 | 南京理工大学 | A kind of multi-connected machine team control type elevator scheduling method |
CN109343532A (en) * | 2018-11-09 | 2019-02-15 | 中国联合网络通信集团有限公司 | A kind of paths planning method and device of dynamic random environment |
CN110065855A (en) * | 2019-04-21 | 2019-07-30 | 苏州科技大学 | Elevator with multiple compartments control method and control system |
CN110127464A (en) * | 2019-05-16 | 2019-08-16 | 永大电梯设备(中国)有限公司 | A kind of multiple target elevator dispatching system and method based on dynamic optimization |
CN114348807A (en) * | 2022-02-15 | 2022-04-15 | 平安科技(深圳)有限公司 | Elevator dispatching method, device, equipment and storage medium |
Family Cites Families (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
FI113467B (en) * | 2002-11-29 | 2004-04-30 | Kone Corp | allocation Method |
NZ536346A (en) * | 2003-11-25 | 2005-11-25 | Inventio Ag | Method of operating a lift installation and lift control |
-
2006
- 2006-05-24 CN CNB200610040554XA patent/CN100413771C/en not_active Expired - Fee Related
Cited By (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108408514A (en) * | 2018-03-14 | 2018-08-17 | 南京理工大学 | A kind of multi-connected machine team control type elevator scheduling method |
CN108408514B (en) * | 2018-03-14 | 2020-04-21 | 南京理工大学 | Multi-connected machine group control type elevator dispatching method |
CN109343532A (en) * | 2018-11-09 | 2019-02-15 | 中国联合网络通信集团有限公司 | A kind of paths planning method and device of dynamic random environment |
CN110065855A (en) * | 2019-04-21 | 2019-07-30 | 苏州科技大学 | Elevator with multiple compartments control method and control system |
CN110065855B (en) * | 2019-04-21 | 2024-01-23 | 苏州科技大学 | Multi-car elevator control method and control system |
CN110127464A (en) * | 2019-05-16 | 2019-08-16 | 永大电梯设备(中国)有限公司 | A kind of multiple target elevator dispatching system and method based on dynamic optimization |
CN110127464B (en) * | 2019-05-16 | 2021-09-17 | 永大电梯设备(中国)有限公司 | Multi-objective elevator dispatching system and method based on dynamic optimization |
CN114348807A (en) * | 2022-02-15 | 2022-04-15 | 平安科技(深圳)有限公司 | Elevator dispatching method, device, equipment and storage medium |
Also Published As
Publication number | Publication date |
---|---|
CN100413771C (en) | 2008-08-27 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN111753468B (en) | Elevator system self-learning optimal control method and system based on deep reinforcement learning | |
US5354957A (en) | Artificially intelligent traffic modeling and prediction system | |
Cortés et al. | Genetic algorithm for controllers in elevator groups: analysis and simulation during lunchpeak traffic | |
US4760896A (en) | Apparatus for performing group control on elevators | |
CN1857981A (en) | Group control lift dispatching method based on CMAC network | |
CN106365003A (en) | Optimized scheduling method for multi-lift-car elevator cluster | |
CN1181741A (en) | Procedure for allocating landing calls in an elevator group | |
Tartan et al. | Optimization of waiting and journey time in group elevator system using genetic algorithm | |
JP2001310878A (en) | Elevator group optimum supervisory operation method | |
CN110980456B (en) | Elevator group control scheduling method based on traffic flow and adaptive neural fuzzy inference | |
CN113741199A (en) | Vehicle economy speed planning method based on intelligent network connection information | |
CN111291921A (en) | Group control elevator dispatching optimization method based on improved artificial bee colony algorithm | |
EP1549581B1 (en) | Elevator group control method | |
CN110689155A (en) | Multi-constraint scheduling method of card collection reservation system considering congestion and emission | |
Yang et al. | Dynamic Partition of Elevator Group Control System with Destination Floor Guidance in Up-peak Traffic. | |
CN113391908B (en) | Reasoning task scheduling method aiming at time delay optimization in vehicle-mounted edge calculation | |
CN115215171A (en) | Multi-car elevator dispatching method based on particle swarm algorithm | |
Liu et al. | Dispatching algorithm design for elevator group control system with Q-learning based on a recurrent neural network | |
Yu et al. | Multi-car elevator system using genetic network programming for high-rise building | |
CN110127464B (en) | Multi-objective elevator dispatching system and method based on dynamic optimization | |
Chen et al. | GA based hybrid fuzzy rule optimization approach for elevator group control system | |
JPH0428681A (en) | Group management control for elevator | |
Chen et al. | Design and implementation of modern elevator group control system | |
Cao et al. | Elevator group dynamic dispatching system based on artificial intelligent theory | |
JPH05246633A (en) | Construction of elevator group control device |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C14 | Grant of patent or utility model | ||
GR01 | Patent grant | ||
C17 | Cessation of patent right | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20080827 Termination date: 20100524 |