CN112811273A - Probability parallel planning based elevator scheduling method based on real-time additivity heuristic method - Google Patents
Probability parallel planning based elevator scheduling method based on real-time additivity heuristic method Download PDFInfo
- Publication number
- CN112811273A CN112811273A CN202110111498.9A CN202110111498A CN112811273A CN 112811273 A CN112811273 A CN 112811273A CN 202110111498 A CN202110111498 A CN 202110111498A CN 112811273 A CN112811273 A CN 112811273A
- Authority
- CN
- China
- Prior art keywords
- heuristic
- additivity
- state
- action set
- reward
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 53
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 53
- 239000000654 additive Substances 0.000 claims description 22
- 230000000996 additive effect Effects 0.000 claims description 21
- 230000033001 locomotion Effects 0.000 claims description 13
- 238000000611 regression analysis Methods 0.000 claims description 11
- 238000007405 data analysis Methods 0.000 claims description 5
- 230000001174 ascending effect Effects 0.000 claims description 3
- 238000013135 deep learning Methods 0.000 claims description 3
- 230000009191 jumping Effects 0.000 claims description 3
- 238000012544 monitoring process Methods 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000009429 distress Effects 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
Images
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B66—HOISTING; LIFTING; HAULING
- B66B—ELEVATORS; ESCALATORS OR MOVING WALKWAYS
- B66B1/00—Control systems of elevators in general
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B66—HOISTING; LIFTING; HAULING
- B66B—ELEVATORS; ESCALATORS OR MOVING WALKWAYS
- B66B1/00—Control systems of elevators in general
- B66B1/34—Details, e.g. call counting devices, data transmission from car to control system, devices giving information to the control system
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B66—HOISTING; LIFTING; HAULING
- B66B—ELEVATORS; ESCALATORS OR MOVING WALKWAYS
- B66B2201/00—Aspects of control systems of elevators
- B66B2201/30—Details of the elevator system configuration
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B66—HOISTING; LIFTING; HAULING
- B66B—ELEVATORS; ESCALATORS OR MOVING WALKWAYS
- B66B2201/00—Aspects of control systems of elevators
- B66B2201/40—Details of the change of control mode
- B66B2201/403—Details of the change of control mode by real-time traffic data
Abstract
The invention discloses a probability parallel programming-based real-time additivity heuristic elevator scheduling method, which abstracts elevator scheduling problems into the field of parallel probability programming and obtains scheduling results through a probability parallel programming-based real-time additivity heuristic algorithm. The invention can efficiently realize elevator dispatching.
Description
Technical Field
The invention relates to the technical field of intelligent planning, in particular to an elevator scheduling method based on probability parallel planning and a real-time additivity heuristic method.
Background
Each elevator has a number to facilitate monitoring and maintenance. Each elevator is provided with a real-time monitor which is responsible for monitoring the up and down of the elevator and sending signals for starting, braking, accelerating, decelerating and opening and closing the elevator door to the elevator lifting box. If the elevator is in failure, a distress signal is also sent to the corresponding elevator responsible person.
In a high-rise building with a large number of elevators and complex passenger demands, the current scheduling algorithm is not efficient, and the problems of no load and long-time waiting of passengers often occur.
Therefore, an efficient elevator dispatching method is urgently needed.
Disclosure of Invention
The invention aims to overcome the defects of the prior art and provide an efficient elevator dispatching method based on probability parallel programming real-time additivity heuristic.
In order to achieve the purpose, the technical scheme provided by the invention is as follows:
a probability parallel programming real-time additivity heuristic elevator scheduling method abstracts elevator scheduling problems into the field of parallel probability programming, and obtains scheduling results through a probability parallel programming real-time additivity heuristic algorithm.
Further, the abstracting of the elevator scheduling problem into the field of parallel probabilistic planning specifically includes:
the object includes: elevators and floors;
the parameters include: the probability parameter of the passenger expecting to reach a certain floor is obtained by counting the frequency;
the state flow includes: the distribution of passengers, the state of the elevator;
the actions include: the elevator moves up and down, and the door is opened and closed;
the reward is as follows: the passengers are correctly transported to the destination, the more passengers are correctly transported, the higher the reward.
Further, the specific process of obtaining the scheduling result by the real-time additive heuristic algorithm based on the probability parallel programming is as follows:
s1, inputting initial state x0Total unit time h;
s2, initializing an additivity heuristic list;
s3, setting h1H/phase _ ratio, phase _ ratio being a phasing superparameter; motion set size estimation vector si=0,i∈[0,h1]If the trace tr is empty, the state count nr is 0;
s4, judging whether nr is less than or equal to h1If yes, acquiring the state xnrThe set of all available actions in the set is called action set AnrOtherwise, go to step S7;
s5, calculating x through action set size estimation algorithmnrBest action set Abest,SnrOptimal action set A calculated for action set size estimation algorithmbestSize of (2), followed by AbestX is to benrEvolution to the next state xnr+1And a reward r to be paid for,adding the data into a history database and a track tr;
s6, updating the heuristic list through an additive numerical value regression analysis method, and returning to the step S4;
s7, predicting the optimal action set size ═ Avg (Σ S)i),i∈[0,h1];
S8, judging whether nr is less than or equal to h, if yes, acquiring the state xnrSet of all available actions in Anr(ii) a Otherwise, outputting the trace tr;
s9, calculating the optimal action set A by the optimal action set algorithmbest;
S10, use AbestX is to benrEvolution to the next state xnr+1And a reward r, which is added to the history database and the track tr;
and S11, updating the heuristic list through an additive numerical value regression analysis method, and returning to the step S8.
Further, the step S2 initializes the additivity heuristic list specifically as follows: in an initial state x0For reference value, all state flow propositions in the field are extracted, and then the initial reward value of the state flow propositions is counted as the value of the item in the initial additivity heuristic list.
Further, the motion set size estimation algorithm compiles the motions including ascending, opening and closing of the elevator into a set, tries are carried out by using different numbers of motions, obtains a motion set with the maximum reward value as a state result, and estimates the optimal motion set size according to the algorithm results of a plurality of states; the specific process is as follows:
a1, input state xnrAnd action set Anr, Ai∈Anr,i∈[0,n]N is the size of the action set Anr;
a2, determining whether i is less than n, if yes, using action AiEvolution State xnrTo xnr'; otherwise, action AiAccording to corresponding xiSorting in descending order to obtain action group Asorted(ii) a Then jumping to step A4;
a3, calculating x by using additive heuristic calculation algorithmnr' reward riI ═ i +1, return to step a 2;
a4, initial value, reward r 0, ntry=0;
A5, judging whether incrase is true, if yes, entering the step A6, otherwise, calculating an optimal action set through an optimal action set algorithm;
a6, judgment Aj∈Asorted,j∈[0,ntry]If yes, then action set Atry=U(Aj),j∈[0,ntry]Using AtryEvolution State xnrTo xnr'; otherwise, ntry=ntry+1, return to step a 5;
A8, determining the rewardIf there is an increase, then the optimal action set Abest=AtryOptimum state xbest=xnr', rewardOtherwise, increment is false;
a9, j ═ j +1, return to step a 6.
Further, the specific process of calculating the optimal action set by the optimal action set algorithm is as follows:
b1, input State xnrAction set AnrTarget action set size;
b2 enumerating all action sets A of sizenrSubset A ofi;
B3, judging whether enumeration is completed or not, if yes, outputting AbestIs the best action set; otherwise, use AiX is to benrEvolution to xnr′;
B5, determining the best reward rbestWhether or not less than or equal toIf so, then Abest=Ai,xbest=xnr',Returning to the step B2; otherwise, go directly back to step B2.
Further, the specific process of the additive heuristic calculation algorithm is as follows:
c1, input State xnrThe reward r is 0;
c2 enumerating all state streams x in state xi;
C3, judging whether enumeration is completed or not, if yes, outputting r as a heuristic value; otherwise, r is increased by x in the heuristic listiCorresponding value and returns to step C2.
Further, the additive numerical regression analysis method updates the heuristic list through a data analysis method.
Further, the data analysis method includes multiple regression or deep learning.
Compared with the prior art, the principle and the advantages of the scheme are as follows:
the elevator scheduling problem is abstracted into the field of parallel probability planning, and the scheduling result is obtained through a real-time additive heuristic algorithm based on the probability parallel planning. The scheme can efficiently realize elevator dispatching.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the services required for the embodiments or the technical solutions in the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to these drawings without creative efforts.
FIG. 1 is a schematic flow chart of a probability-based parallel programming real-time additivity heuristic algorithm (Algorithm 1) according to an embodiment of the present invention;
FIG. 2 is a schematic flow chart of an initialization algorithm (Algorithm 2) in an embodiment of the present invention;
FIG. 3 is a schematic flow chart of an action set size estimation algorithm (Algorithm 3) in an embodiment of the present invention;
FIG. 4 is a schematic flow chart of the optimal action set algorithm (Algorithm 4) in an embodiment of the present invention;
FIG. 5 is a schematic flow chart of an additive heuristic calculation algorithm (Algorithm 5) in an embodiment of the present invention;
FIG. 6 is a schematic flow chart of an additive numerical regression analysis method (Algorithm 6) in an embodiment of the present invention.
Detailed Description
The invention will be further illustrated with reference to specific examples:
the elevator scheduling method based on the probability parallel programming real-time additivity heuristic method abstracts elevator scheduling problems into the field of parallel probability programming, and obtains scheduling results through the probability parallel programming real-time additivity heuristic algorithm.
Wherein, abstract the elevator scheduling problem for parallel probability planning field specifically includes:
the object includes: elevators and floors;
the parameters include: the probability parameter of the passenger expecting to reach a certain floor is obtained by counting the frequency;
the state flow includes: the distribution of passengers, the state of the elevator;
the actions include: the elevator moves up and down, and the door is opened and closed;
the reward is as follows: the passengers are correctly transported to the destination, the more passengers are correctly transported, the higher the reward.
As shown in fig. 1, the specific process of obtaining the scheduling result by the heuristic algorithm based on the probability parallel programming real-time additivity is as follows:
s1, inputting initial state x0Total unit time h;
s2, initializing an additivity heuristic list ahl through an initialization algorithm, specifically: in an initial state x0Extracting all state flow propositions in the field as reference values, and then counting the initial reward value of the proposition as the value of the item in the initial additivity heuristic list, as shown in FIG. 2;
s3, setting h1H/phase _ ratio, phase _ ratio being a phasing superparameter; motion set size estimation vector si=0,i∈[0,h1]If the trace tr is empty, the state count nr is 0;
s4, judging whether nr is less than or equal to h1If yes, acquiring the state xnrThe set of all available actions in the set is called action set AnrOtherwise, go to step S7;
s5, calculating x through action set size estimation algorithmnrBest action set Abest,
In the step, the motion set size estimation algorithm compiles the motions including ascending, opening and closing of the elevator into a set, tries are carried out by using different numbers of motions, obtains the motion set with the maximum reward value as the result of the state, and estimates the size of the optimal motion set according to the algorithm results of a plurality of states; as shown in fig. 3, the specific process is as follows:
a1, input state xnrAnd action set Anr, Ai∈Anr,i∈[0,n]N is the size of the action set Anr;
a2, determining whether i is less than n, if yes, using action AiEvolution State xnrTo xnr'; otherwise, action AiAccording to corresponding xiSorting in descending order to obtain action group Asorted(ii) a Then jumping to step A4;
a3, calculating x by using additive heuristic calculation algorithmnr' reward riI ═ i +1, return to step a 2;
a4, initial value, reward r 0, ntry=0;
A5, judging whether incrase is true, if yes, entering the step A6, otherwise, calculating an optimal action set through an optimal action set algorithm;
a6, judgment Aj∈Asorted,j∈[0,ntry]If yes, then action set Atry=U(Aj),j∈[0,ntry](U represents union), using AtryEvolution State xnrTo xnr'; otherwise, ntry=ntry+1, return to step a 5;
A8, determining the rewardIf there is an increase, then the optimal action set Abest=AtryOptimum state xbest=xnr', rewardOtherwise, increment is false;
a9, j ═ j +1, return to step a 6.
SnrOptimal action set A calculated for action set size estimation algorithmbestSize of (2), followed by AbestX is to benrEvolution to the next state xnr+1And a reward r, which is added to the history database and the track tr;
s6, updating the heuristic list (as shown in FIG. 6) by an additive numerical regression analysis method, and returning to the step S4;
s7, predicting the optimal action set size ═ Avg (Σ S)i),i∈[0,h1];
S8, judging whether nr is less than or equal to h, if yes, acquiring the state xnrSet of all available actions in Anr(ii) a Otherwise, outputting the trace tr;
S9calculating an optimal action set A through an optimal action set algorithmbestAs shown in fig. 4, the specific process includes:
b1, input State xnrAction set AnrTarget action set size;
b2 enumerating all action sets A of sizenrSubset A ofi;
B3, judging whether enumeration is completed or not, if yes, outputting AbestIs the best action set; otherwise, use AiX is to benrEvolution to xnr′;
B4 calculating x by additive heuristic calculation algorithmnr' s rewardAs shown in fig. 5, the process includes:
c1, input State xnrThe reward r is 0;
c2 enumerating all state streams x in state xi;
C3, judging whether enumeration is completed or not, if yes, outputting r as a heuristic value; otherwise, r is increased by x in the heuristic listiCorresponding value and returns to step C2.
B5, determining the best reward rbestWhether or not less than or equal toIf so, then Abest=Ai,xbest=xnr',Returning to the step B2; otherwise, go directly back to step B2.
S10, use AbestX is to benrEvolution to the next state xnr+1And a reward r, which is added to the history database and the track tr;
s11, updating the heuristic list through an additive numerical regression analysis method (as shown in FIG. 6), and returning to the step S8.
In this embodiment, the heuristic list is updated by an additive numerical regression analysis method through a data analysis method, which includes, but is not limited to, deep learning, multiple regression analysis, and the like.
The above-mentioned embodiments are merely preferred embodiments of the present invention, and the scope of the present invention is not limited thereto, so that variations based on the shape and principle of the present invention should be covered within the scope of the present invention.
Claims (9)
1. A probability parallel programming based real-time additivity heuristic elevator scheduling method is characterized in that elevator scheduling problems are abstracted into the field of parallel probability programming, and scheduling results are obtained through a probability parallel programming based real-time additivity heuristic algorithm.
2. The elevator scheduling method based on the probabilistic parallel programming real-time additivity heuristic method of claim 1, wherein abstracting the elevator scheduling problem into the field of parallel probabilistic programming specifically comprises:
the object includes: elevators and floors;
the parameters include: the probability parameter of the passenger expecting to reach a certain floor is obtained by counting the frequency;
the state flow includes: the distribution of passengers, the state of the elevator;
the actions include: the elevator moves up and down, and the door is opened and closed;
the reward is as follows: the passengers are correctly transported to the destination, the more passengers are correctly transported, the higher the reward.
3. The elevator scheduling method according to claim 2, wherein the specific process of obtaining the scheduling result by the probabilistic parallel programming real-time additivity heuristic based on the probabilistic parallel programming real-time additivity heuristic algorithm is as follows:
s1, inputting initial state x0Total unit time h;
s2, initializing an additivity heuristic list ahl through an initialization algorithm;
s3, setting h1H/phase _ ratio, phase _ ratio being a phasing superparameter; motion set size estimation vector si=0,i∈[0,h1]If the trace tr is empty, the state count nr is 0;
s4, judging whether nr is less than or equal to h1If yes, acquiring the state xnrThe set of all available actions in the set is called action set AnrOtherwise, go to step S7;
s5, calculating x through action set size estimation algorithmnrBest action set Abest,SnrOptimal action set A calculated for action set size estimation algorithmbestSize of (2), followed by AbestX is to benrEvolution to the next state xnr+1And a reward r, which is added to the history database and the track tr;
s6, updating the heuristic list through an additive numerical value regression analysis method, and returning to the step S4;
s7, predicting the optimal action set size ═ Avg (Σ S)i),i∈[0,h1];
S8, judging whether nr is less than or equal to h, if yes, acquiring the state xnrSet of all available actions in Anr(ii) a Otherwise, outputting the trace tr;
s9, calculating the optimal action set A by the optimal action set algorithmbest;
S10, use AbestX is to benrEvolution to the next state xnr+1And a reward r, which is added to the history database and the track tr;
and S11, updating the heuristic list through an additive numerical value regression analysis method, and returning to the step S8.
4. The elevator dispatching method based on the probabilistic parallel programming real-time additivity heuristic, according to claim 3, wherein the step S2 initializes the additivity heuristic list ahl specifically as: in an initial state x0For reference value, all state flow propositions in the field are extracted, and then the initial reward value of the state flow propositions is counted as the value of the item in the initial additivity heuristic list.
5. The elevator dispatching method based on the probability parallel programming real-time additivity heuristic method of claim 3, characterized in that the action set size estimation algorithm compiles actions including ascending, opening and closing of an elevator into a set, tries by using different numbers of the actions, obtains an action set with the largest reward value as a result of a state, and estimates the size of the optimal action set by the algorithm results of a plurality of states; the specific process is as follows:
a1, input state xnrAnd action set Anr, Ai∈Anr,i∈[0,n]N is the size of the action set Anr;
a2, determining whether i is less than n, if yes, using action AiEvolution State xnrTo xnr'; otherwise, action AiAccording to corresponding xiSorting in descending order to obtain action group Asorted(ii) a Then jumping to step A4;
a3, calculating x by using additive heuristic calculation algorithmnr' reward riI ═ i +1, return to step a 2;
a4, initial value, reward r 0, ntry=0;
A5, judging whether incrase is true, if yes, entering the step A6, otherwise, calculating an optimal action set through an optimal action set algorithm;
a6, judgment Aj∈Asorted,j∈[0,ntry]If yes, then action set Atry=U(Aj),j∈[0,ntry]Using AtryEvolution State xnrTo xnr'; otherwise, ntry=ntry+1, return to step a 5;
A8, determining the rewardIf there is an increase, then the optimal action set Abest=AtryOptimum state xbest=xnr', rewardOtherwise, increment is false;
a9, j ═ j +1, return to step a 6.
6. The elevator dispatching method based on the probabilistic parallel programming real-time additivity heuristic method of claim 5, wherein the specific process of calculating the optimal action set through the optimal action set algorithm is as follows:
b1, input State xnrAction set AnrTarget action set size;
b2 enumerating all action sets A of sizenrSubset A ofi;
B3, judging whether enumeration is completed or not, if yes, outputting AbestIs the best action set; otherwise, use AiX is to benrEvolution to xnr’;
7. The elevator dispatching method based on the probability parallel programming real-time additivity heuristic method of claim 6, wherein the concrete process of the additivity heuristic calculation algorithm is as follows:
c1, input State xnrThe reward r is 0;
c2 enumerating all state streams x in state xi;
C3, judging whether enumeration is completed or not, if yes, outputting r as a heuristic value; otherwise, r is increased by x in the heuristic listiCorresponding value and returns to step C2.
8. The elevator dispatching method based on the probabilistic parallel programming real-time additivity heuristic method of claim 3, wherein the additivity numerical regression analysis method updates the heuristic list through a data analysis method.
9. The elevator dispatching method based on the probabilistic parallel programming real-time additivity heuristic method of claim 8, wherein the data analysis comprises multivariate regression or deep learning.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110111498.9A CN112811273A (en) | 2021-01-27 | 2021-01-27 | Probability parallel planning based elevator scheduling method based on real-time additivity heuristic method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110111498.9A CN112811273A (en) | 2021-01-27 | 2021-01-27 | Probability parallel planning based elevator scheduling method based on real-time additivity heuristic method |
Publications (1)
Publication Number | Publication Date |
---|---|
CN112811273A true CN112811273A (en) | 2021-05-18 |
Family
ID=75859826
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110111498.9A Pending CN112811273A (en) | 2021-01-27 | 2021-01-27 | Probability parallel planning based elevator scheduling method based on real-time additivity heuristic method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN112811273A (en) |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106203634A (en) * | 2016-07-20 | 2016-12-07 | 广东工业大学 | A kind of based on the didactic parallel probability plan method of cause-and-effect diagram |
US20180204111A1 (en) * | 2013-02-28 | 2018-07-19 | Z Advanced Computing, Inc. | System and Method for Extremely Efficient Image and Pattern Recognition and Artificial Intelligence Platform |
CN109002914A (en) * | 2018-07-11 | 2018-12-14 | 广东工业大学 | A kind of production scheduling method and device merging random algorithm and heuristic programming |
CN110562821A (en) * | 2019-09-29 | 2019-12-13 | 中冶长天国际工程有限责任公司 | Mine hoist falling self-rescue system and control method |
-
2021
- 2021-01-27 CN CN202110111498.9A patent/CN112811273A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20180204111A1 (en) * | 2013-02-28 | 2018-07-19 | Z Advanced Computing, Inc. | System and Method for Extremely Efficient Image and Pattern Recognition and Artificial Intelligence Platform |
CN106203634A (en) * | 2016-07-20 | 2016-12-07 | 广东工业大学 | A kind of based on the didactic parallel probability plan method of cause-and-effect diagram |
CN109002914A (en) * | 2018-07-11 | 2018-12-14 | 广东工业大学 | A kind of production scheduling method and device merging random algorithm and heuristic programming |
CN110562821A (en) * | 2019-09-29 | 2019-12-13 | 中冶长天国际工程有限责任公司 | Mine hoist falling self-rescue system and control method |
Non-Patent Citations (1)
Title |
---|
饶东宁: "基于因果图启发式的并行概率规划求解", 《计算机应用研究》 * |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
EP0565864B1 (en) | Artificially intelligent traffic modelling and prediction system | |
Hirasawa et al. | A double-deck elevator group supervisory control system using genetic network programming | |
CN110171753A (en) | A kind of elevator dispatching strategy processing method, device, equipment and storage medium | |
JPH10194611A (en) | Group supervisory operation control method of elevator | |
CN106586738A (en) | Multi-car elevator dispatching method based on visual inspection | |
CN114330138A (en) | Fault diagnosis method and device and electronic equipment | |
JPH03279178A (en) | Control device for elevator | |
CN112811273A (en) | Probability parallel planning based elevator scheduling method based on real-time additivity heuristic method | |
Jan et al. | Hybrid stochastic ranking for constrained optimization | |
CN114387024A (en) | User analysis system for E-commerce repurchase behavior based on Mamdani algorithm | |
Yu et al. | Multicar elevator group supervisory control system using genetic network programming | |
CN115072508A (en) | Efficient dispatching method for elevators in intelligent building | |
CN116001864A (en) | Automatic machine and reinforcement learning-based railway traffic system scheduling optimization method | |
Beamurgia et al. | Improving waiting time and energy consumption performance of a bi-objective genetic algorithm embedded in an elevator group control system through passenger flow estimation | |
Rashid et al. | Design and implementation of fuzzy based controller for modern elevator group | |
JP6776174B2 (en) | Elevator user movement prediction method and elevator user movement prediction device | |
CN114814973A (en) | Intelligent security check system and method for man-machine hybrid decision | |
CN114707587A (en) | Traffic pattern recognition method based on genetic algorithm and fuzzy neural network | |
EP3620416B1 (en) | Elevator system management utilizing machine learning | |
CN110790096A (en) | Elevator stopping strategy making method, elevator control method and device and elevator system | |
Wan et al. | Traffic pattern-aware elevator dispatching via deep reinforcement learning | |
Yu et al. | Multi-car elevator system using genetic network programming | |
JP7294480B1 (en) | Learning device, reasoning device and elevator group control device | |
Zhou et al. | Elevator group supervisory control system using genetic network programming with macro nodes and reinforcement learning | |
Inamoto et al. | Model-approximated dynamic programming based on decomposable state transition probabilities |
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 | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20210518 |