CN112960496B

CN112960496B - Intelligent elevator management system and method based on intelligent gateway

Info

Publication number: CN112960496B
Application number: CN202110242545.3A
Authority: CN
Inventors: 徐胜; 刘友华; 戈云娇; 陆勤灵; 李舒帆; 李红雨; 李九生; 刘丽敏
Original assignee: Foshan Sanshui District Polytechnic School Foshan Sanshui District Polytechnic School
Current assignee: Foshan Sanshui District Polytechnic School Foshan Sanshui District Polytechnic School
Priority date: 2021-03-04
Filing date: 2021-03-04
Publication date: 2023-02-07
Anticipated expiration: 2041-03-04
Also published as: CN112960496A

Abstract

The invention relates to an intelligent elevator management system and method based on an intelligent gateway, wherein the system comprises an Internet of things device, a server and the intelligent gateway, and the server is configured as follows: transmitting first information acquired based on reservation information to the Internet of things equipment through the intelligent gateway; planning a navigation path for an appointment person and dynamically adjusting a dispatching/control strategy of an elevator at least based on second information and first positioning information which are uploaded by the Internet of things equipment and related to an action path due to the triggering of the first information; and/or navigating the reservation personnel and dynamically adjusting the dispatching/control strategy of the elevator based on the second positioning information uploaded by the reservation personnel. Through the setting mode, the invention can fuse non-real-time off-line position information and on-line positioning information, and formulate a scheduling/control strategy at two angles of anticipating the arrival time of reservation personnel and reducing the randomness of elevator selection, thereby reducing the unintended operation of the elevator.

Description

Intelligent elevator management system and method based on intelligent gateway

Technical Field

The invention relates to the technical field of elevator internet of things, in particular to an intelligent elevator management system and method based on an intelligent gateway.

Background

With the continuous emergence of high-rise buildings, elevators are used as transportation tools which are important components of intelligent buildings to bear the riding demands of a large number of office workers. On one hand, the elevator is used as a public tool of a personnel gathering place and needs to be ensured to run safely, and on the other hand, the problem of the running efficiency of the elevator is more prominent because a large number of personnel enter and exit frequently and are distributed in a tidal manner. For example, in the peak time of going to work and getting off work, the elevator is not available for a long time, or passes through the floor and stops continuously, or stops layer by layer, so that the operation efficiency of the elevator is low, and people flow cannot be dredged quickly.

For example, chinese patent publication No. CN110342355A discloses an intelligent building elevator dispatching platform system and a working method thereof, the system includes a client, a cloud server, a communication gateway, a data acquisition terminal, a Zigbee coordinator, an intelligent gateway, an intelligent building elevator group control dispatching master control system and a sub-control system; the client is connected with the cloud server through a wireless network; the client side realizes elevator dispatching, elevator reservation in advance, elevator position checking, elevator number monitoring, remote reservation and arrival reminding through installed software; the data acquisition terminal establishes wireless connection through a Zigbee coordinator; the Zigbee coordinator is in wireless connection with the intelligent gateway, and the intelligent gateway is accessed to the Internet sequentially through the router and the TCP/IP network so as to be connected with the cloud server; the cloud server monitors and analyzes the elevator transportation data through big data analysis and cloud computing on the data uploaded by the intelligent gateway, monitors the running conditions of all elevators in real time, and establishes communication connection with an intelligent building elevator group control dispatching master control system through a communication gateway; the intelligent building elevator group dispatching master control system analyzes the optimal stopping and running time, reasonably plans the elevator taking time, realizes the optimized dispatching of the elevator through the branch control system, reduces the energy consumption, realizes the efficient running of the elevator, and improves the safety and the taking experience of people. Although the patent can acquire the target floor of a passenger through functions of reserving an elevator in advance or remote reservation and the like, and establishes connection with sensing equipment for monitoring the running state of the elevator at the bottom layer through technologies such as the Internet of things and the like, the related data of the target floor and the running state data of the elevator are transmitted to the server, and the optimal service path of the elevator is predicted through data analysis and calculation at the server side, so that the stop times of the elevator are reduced, the running efficiency of the elevator is improved, and people flow is dredged quickly. However, in the practice of high-rise building management, besides people may be gathered during the peak hours of work and work, a large number of other people who frequently enter and exit frequently visit office companies in the high-rise buildings every day, and the elevator in different high-rise buildings is different in installation position, different in accessible floor sections, and easily confused by passengers and goods elevators, so that the resident office workers cannot quickly find the elevator which can reach the target floor, and the uncertainty of the arrival time of the visitor and the uncertainty of elevator selection improve the difficulty of the server in calculating and predicting the elevator service path. On the other hand, the uncertainty of the selection of an elevator by a visitor and the uncertainty of the selection of an elevator by an unsubscribed person increase the unintended operation of the elevator, which in turn leads to an increase in energy consumption and wear of the equipment.

Furthermore, on the one hand, due to the differences in understanding to the person skilled in the art; on the other hand, since the inventor has studied a lot of documents and patents when making the present invention, but the space is not limited to the details and contents listed in the above, however, the present invention is by no means free of the features of the prior art, but the present invention has been provided with all the features of the prior art, and the applicant reserves the right to increase the related prior art in the background.

Disclosure of Invention

Aiming at the defects of the prior art, the invention provides an intelligent elevator management system based on an intelligent gateway, which comprises an Internet of things device, a server and the intelligent gateway. The server is configured to: transmitting first information acquired based on reservation information to the Internet of things equipment through the intelligent gateway; planning a navigation path for an appointment person and dynamically adjusting a dispatching/control strategy of an elevator at least based on second information and first positioning information which are uploaded by the Internet of things equipment and related to an action path due to the triggering of the first information; and/or navigating the reservation personnel and dynamically adjusting the dispatching/control strategy of the elevator based on the second positioning information uploaded by the reservation personnel. Preferably, the first information may include a destination floor of the reservation person, physiological information of the reservation person, authority information of the reservation person, identification information of an electronic device of the reservation person, and the like. The physiological information of the reservation personnel comprises face information, fingerprint information and the like. Preferably, the second information includes at least ID information and time information of the internet of things device. The time information is generated by the internet of things device based on the triggering of the first information. For example, when the internet of things device is an access control device, the access control device verifies the identity of the reservation person based on the first information transmitted by the server. The entrance guard equipment can verify the identity of the reservation personnel through face recognition, fingerprint recognition and electronic equipment of the reservation personnel. The identity verification by using the electronic equipment of the reservation personnel can be that the server sends the two-dimensional code corresponding to the first information, when the reservation personnel pass through the access control equipment, the access control equipment scans the two-dimensional code displayed by the electronic equipment of the reservation personnel and then decrypts, and if the decrypted information is consistent with the first information, the identity of the reservation personnel is judged. Preferably, the first information may be a public key. The access control equipment can utilize the first information to decrypt the two-dimensional code displayed by the reservation personnel. Preferably, the server is configured to anticipate an expected arrival time of the reservation person and the selected elevator based at least on the second information and the second positioning message. The server is configured to dynamically adjust a dispatch/control strategy for an elevator based on the expected arrival time and the selected elevator. The existing intelligent scheduling or control strategy for the elevator is based on the states of the elevator and the reservation personnel whether the optimization targets such as energy consumption, elevator traveling distance, elevator taking personnel waiting time and the like are selected or the service functions such as remote reservation, elevator running state providing, safety monitoring and the like are selected, however, the real state of the reservation personnel side, namely the actual arrival time of the reservation personnel at the elevator and the selected elevator door are not considered in the prior art, so that the existing intelligent scheduling or control strategy for the elevator has large uncertainty, and the existing intelligent scheduling or control strategy for the elevator is not in accordance with the actual situation, such as no personnel take the elevator at the appointed time and the unexpected more/less loads of the elevator car are carried by the elevator. All of the above situations can cause unexpected operation of the elevator, thereby increasing energy consumption and equipment wear. According to the invention, the position of the reservation personnel is sensed by fusing the off-line second information and the on-line first positioning information, namely the second information of the reservation personnel is acquired in a non-real-time and off-line manner by sensing the equipment of the Internet of things at a fixed position, the real-time on-line position information of the reservation personnel is acquired by the indoor positioning equipment, and the real-time on-line position information is rectified by the off-line position information fixed by the Internet of things, so that the expected arrival time of the reservation personnel is acquired and the reservation personnel is navigated to reduce the uncertainty of selecting the elevator by the reservation personnel, and further, the server can integrate the arrival time of the reservation personnel and the certainty of selecting the elevator when a scheduling/control strategy of the elevator is formulated, so that the generated scheduling/control strategy can reduce the unexpected operation of the elevator, and further avoid the increase of energy consumption and equipment abrasion.

The invention also provides an intelligent elevator management system based on the intelligent gateway, which comprises the Internet of things equipment, the server and the intelligent gateway. The intelligent gateway is configured to:

requesting, in a remote configuration of the server and/or the internet of things devices, the internet of things devices to feed back second information and first positioning information about an action path generated based on the reserved first information in a manner of periodically accessing the plurality of internet of things devices, and transmitting the second information and the first positioning information to the server;

in the process of periodic access, at least two pieces of Internet of things equipment which generate the first positioning messages and have the signal strength difference exceeding a first threshold value are searched in a one-by-one comparison mode on the basis of the Internet of things equipment which sends the second information.

The invention also provides an intelligent elevator management system based on the intelligent gateway, which comprises the Internet of things equipment and the intelligent gateway. The Internet of things equipment is configured as follows:

and feeding back second information about offline paths of reserved people to the intelligent gateway based on the first information about reservation transmitted by the intelligent gateway. And issuing a radio frequency signal and/or an optical signal under the instruction of the intelligent gateway and/or the triggering of the second information so as to acquire the radio frequency signal and/or the optical signal which are fed back by the electronic equipment of the reservation personnel and are related to the online position, and transmitting the fed back radio frequency signal and/or the fed back optical signal to the intelligent gateway.

According to a preferred embodiment, the server is configured to anticipate the dispatch/control strategy of an elevator as follows:

acquiring first information about a reservation;

predicting elevator traffic demands based on the real-time elevator running state and the historical elevator running state fed back by the Internet of things equipment;

a dispatch/control strategy for the elevator is anticipated based on the first information and the elevator traffic demand.

According to a preferred embodiment, the server is configured to obtain the dispatch/control strategy of the elevator based on a deep reinforcement learning/reinforcement learning algorithm. The optimization target is as follows: the waiting time is shortest and the energy consumption is lowest; the constraint conditions are as follows: the elevator operation efficiency is not reduced, and the load difference of the elevator cars is smaller than a third threshold value; the strategy variables are: the number of stops of the elevator and the travel distance. Adding a loss function with respect to latency characterized by the first positioning message/the second positioning message in the deep reinforcement learning/reinforcement learning update function.

According to a preferred embodiment, the intelligent gateway is configured to merge the second information and the first positioning message according to the following steps:

and searching at least two pieces of Internet of things equipment which generate the first positioning message and have the signal strength difference of the first positioning message exceeding a first threshold value in a one-by-one comparison mode based on the Internet of things equipment which sends the second information. Preferably, starting with the physical location of the internet of things device which first sends the second information, the internet of things devices are requested to feed back the first positioning messages one by one according to the pre-stored IDs of the internet of things devices which are sorted by the actual physical location. And in the process of searching at least two pieces of Internet of things equipment which generate the first positioning messages and have the signal difference of the first positioning messages exceeding a first threshold value, periodically requesting the Internet of things equipment to feed back second information. And if the second information is acquired, adjusting the currently acquired first positioning message based on the second information to avoid the position information transmitted by the first positioning message from deviating from the actual position.

According to a preferred embodiment, in the case that the difference between the signal strengths of at least two first positioning messages exceeds the second threshold value in a plurality of internet of things devices transmitting the first positioning messages, the intelligent gateway is configured to:

shielding the Internet of things equipment of which the signal intensity of the first positioning message is smaller than a fourth threshold value, and requesting to close the Internet of things equipment;

and searching the IOT equipment which does not transmit the first positioning message by taking the ID of the IOT equipment as an initial point, so that the intelligent gateway can keep communication with at least two IOT equipment which transmit the first positioning message and have the signal strength difference of the first positioning message exceeding a first threshold value.

According to a preferred embodiment, in the case that, among a plurality of internet of things devices transmitting first positioning messages, a difference between signal strengths of at least two first positioning messages exceeds a second threshold value and there is no internet of things device transmitting a first positioning message whose signal strength is smaller than a fourth threshold value, the smart gateway is configured to:

shielding the Internet of things equipment of which the signal intensity of the first positioning message is greater than a fifth threshold value, and requesting to close the Internet of things equipment;

and searching the Internet of things equipment which does not transmit the first positioning message by taking the ID of the Internet of things equipment as an initial point, so that the intelligent gateway can keep communication with at least two pieces of Internet of things equipment which transmit the first positioning message and have the signal intensity difference of the first positioning message exceeding a first threshold value.

The invention also provides an intelligent elevator management method based on the intelligent gateway, which comprises the following steps:

transmitting first information acquired based on reservation information to the Internet of things equipment through the intelligent gateway;

planning a navigation path for a reservation staff and dynamically adjusting a dispatching/control strategy of an elevator at least based on second information and first positioning information which are uploaded by the Internet of things equipment and related to the action path due to the triggering of the first information; and/or

And navigating the reservation personnel and dynamically adjusting the dispatching/control strategy of the elevator based on the second positioning message uploaded by the reservation personnel.

acquiring first information about a reservation;

a dispatch/control strategy for an elevator is anticipated based on the first information and elevator traffic demand.

Drawings

Fig. 1 is a simplified block diagram of a preferred embodiment of the intelligent elevator management system of the present invention;

fig. 2 is a schematic flow chart showing the steps of a preferred embodiment of the intelligent elevator management method of the present invention.

List of reference numerals

100: the internet of things device 200: the server 300: intelligent gateway

Detailed Description

The following detailed description is made with reference to the accompanying drawings.

The elevator Internet of things is a concept provided for solving the safety problem of the existing elevator, and a data acquisition part, a data transmission part, a central processing part and application software jointly form a complete elevator Internet of things monitoring system. The collecting instrument collects elevator operation data, analyzes the elevator operation data, uploads the elevator operation data to the internet monitoring center, and combines platform application software, so that real-time and effective supervision and maintenance of the elevator by each relevant unit are realized. However, referring to the intelligent processes of other industries, some problems existing in the intelligent processes can be found, one of the obvious characteristics is that communication interfaces and protocols of a plurality of sensing layers are different and uniform, sensing devices with different interfaces are difficult to access to an internet system, and along with the development of the internet of things, a large number of new technologies appear in each layer of sensing, transmission and application, so that different technical schemes may be adopted, if the sensing devices are respectively accessed to a network, the construction cost can be greatly increased, and the problem exists in each intelligent subsystem. If the Zigbee and 485 buses are adopted in the intelligent home, the communication type and the protocol of the intelligent bulb may be different from those of the intelligent socket. Different communication technologies such as CAN, 485, zigbee and 232 are also used in the elevator, each communication interface CAN generate data, along with the development of a smart city, more elevator information CAN be uploaded to a network platform, and interfaces are opened to government regulatory departments. It is unlikely that the interfaces between different industries are unified in a short period, but the industry convergence is a trend, so that the adoption of the intelligent gateway 300 is a more probable scheme in the elevator internet of things and elevator intelligent process. The intelligent gateway 300, as a gateway device, can implement protocol conversion between the sensing network and the communication network and between different types of network sensing networks, and can implement both time-wide area interconnection and local area interconnection. In addition, the intelligent gateway 300 also has an equipment management function, and an operator can manage all sensing nodes on the bottom layer through the intelligent gateway 300, know the relevant information of the nodes and realize remote control, so that the integration of different sensing layers of the Internet of things of the elevator can be realized at lower cost.

Preferably, the intelligent gateway 300 mainly functions to implement protocol conversion between different types of networks, i.e., repackaging data packets transmitted by different networks to conform to the communication protocol of the network to be connected. Preferably, the intelligent gateway 300 has a plurality of reusable extension interfaces, including at least a 2G/3G/4G/5G interface and a local Ethernet interface.

The internet of things device 100 may be an access control device, an indoor positioning device, or a monitoring device. The internet of things device 100 may also be a device capable of collecting an operation state of the elevator. For example, the internet of things device 100 may be an infrared sensing sensor, a wireless positioning sensor, an in-elevator video acquisition device, an in-elevator communication device, an elevator fault acquisition device, and the like.

Preferably, the server 200 may be a remote server, and may also be a cloud server. Preferably, the server 200 may be used for cloud computing as well as cloud storage.

Preferably, the electronic device of the reservation person may be a computer device, such as a mobile computing device, a notebook, a tablet, a mobile phone. The electronic device of the reservation person may also be a smart wearable device, such as a smart watch, smart glasses, or the like.

Preferably, the electronic device of the reservation person can access to a network such as the internet, the internet of things, a mobile network and the like in a wired or wireless manner.

Example 1

As shown in fig. 1, the invention provides an intelligent elevator management system based on an intelligent gateway, which includes an internet of things device 100, a server 200 and an intelligent gateway 300. The server 200 is configured to: transmitting first information acquired based on the reservation information to the internet of things device 100 through the intelligent gateway 300; planning a navigation path for a reservation staff and dynamically adjusting a dispatching/control strategy of an elevator at least based on second information and first positioning information which are uploaded by the Internet of things equipment 100 and related to the action path and are triggered by the first information; and/or navigating the reservation personnel and dynamically adjusting the dispatching/control strategy of the elevator based on the second positioning information uploaded by the reservation personnel. Preferably, the first information may include a destination floor of the reservation person, physiological information of the reservation person, authority information of the reservation person, identification information of an electronic device of the reservation person, and the like. The physiological information of the reservation personnel comprises face information, fingerprint information and the like. Preferably, the second information includes at least ID information and time information of the internet of things device 100. The time information is time information generated by the internet of things device 100 based on the triggering of the first information. Preferably, the server 200 may obtain the location information through ID information of the internet of things device 100. Since the location of the internet of things device 100 is fixed, the location information of the internet of things device 100 can be mapped one to one by compiling an ID in advance. For example, when the internet of things device 100 is an access control device, the access control device verifies the identity of a reservation person based on the first information transmitted from the server 200. The entrance guard equipment can verify the identity of the reservation personnel through face recognition, fingerprint recognition and electronic equipment of the reservation personnel. The identity verification by using the electronic equipment of the reservation personnel can be that the server 200 sends the two-dimensional code corresponding to the first information, when the reservation personnel pass through the access control equipment, the access control equipment scans the two-dimensional code displayed by the electronic equipment of the reservation personnel and then decrypts, and if the decrypted information is consistent with the first information, the identity of the reservation personnel is judged. Preferably, the first information may be a public key. The access control equipment can utilize the first information to decrypt the two-dimensional code displayed by the reservation personnel. Preferably, the server 200 is configured to anticipate the expected arrival time of the reservation person and the selected elevator based at least on said second information and second positioning message. The server 200 is configured to dynamically adjust the dispatch/control strategy of the elevator based on the expected arrival time and the selected elevator. Preferably, the second positioning message may be a manner in which a reservation person may connect to the server 200 through the internet of things device 100 or an intelligent gateway, and actively upload the location information thereof. Through the above setting mode, the beneficial effect who reaches is:

the prior art does not consider the relation between an elevator and position navigation, on one hand, the position setting or elevator planning of the elevator is not standard, and a reservation person may not find the elevator or the selected elevator cannot reach the target floor, so the position navigation function is needed. On the other hand, the intelligent dispatching and control of the elevator not only needs to monitor the running state of the elevator, but also needs to make an appointment for the exact arrival time of the personnel for planning, thereby reducing the operation of unexpected plans. That is, the running state of the elevator itself is changed from moment to moment, and if the time exactly reached by the reservation personnel is not available, the currently established or expected elevator service path is not in accordance with the reality, thereby causing the unexpected running of the elevator, increasing the abrasion of equipment and the consumption of energy. The elevator management system of the present invention uses the internet of things of the elevator, i.e., the server 200 acquires the location information and time information of the reservation person in non-real time and offline by sensing the fixed internet of things device 100. For example, when a reservation person passes through the access control device, the access control device allows the reservation person to pass through based on first information of the reservation person, and feeds back the time when the reservation person passes through the access control device and the ID of the reservation person to the server 200, so that the server 200 can know the action track of the reservation person. The server 200 can quickly plan the navigation path of the reservation staff and guide the reservation staff to reach the designated elevator. Meanwhile, the action track of the reservation personnel acquired through the internet of things equipment 100 similar to the access control equipment is non-real-time offline position information, on one hand, the navigation path planned by the server 200 may deviate from the actual action path of the reservation personnel due to the hysteresis of acquiring the position information of the reservation personnel, on the other hand, the server 200 cannot correct the deviation of the reservation personnel when the reservation personnel deviates from the navigation path or does not select a corresponding elevator because the position information of the reservation personnel cannot be acquired in real time, and on the other hand, the planned elevator dispatching scheme or control scheme cannot be adjusted and corrected in real time, so that a large amount of unexpected operation of the elevator is caused. Therefore, the invention acquires the first positioning message or the second positioning message by utilizing the indoor positioning function in the internet of things device 100 to acquire the real-time online position information of the reservation personnel, and corrects the real-time online position information by the offline position information fixed by the internet of things, so as to acquire the expected arrival time of the reservation personnel and navigate for the reservation personnel to reduce the uncertainty of selecting the elevator for the reservation personnel, and further the server 200 can integrate the arrival time of the reservation personnel and the certainty of selecting the elevator when determining the scheduling/control strategy of the elevator, so that the generated scheduling/control strategy can reduce the unexpected operation of the elevator, and further avoid the increase of energy consumption and the increase of equipment wear.

Preferably, the server 200 is configured to anticipate the dispatch/control strategy of the elevators as follows:

acquiring first information about a reservation;

predicting elevator traffic demands based on real-time elevator running states and historical elevator running states fed back by the Internet of things equipment 100;

a dispatch/control strategy for the elevator is anticipated based on the first information and the elevator traffic demand. Preferably, the first information comprises a destination address provided by the reservation operator. The destination address includes a destination floor to which the reservation person is to reach. Preferably, the server 200 may employ a fuzzy algorithm for logic processing to sense peak elevator traffic. The server 200 may also be configured to obtain the demand for elevator traffic based on a risk prediction of elevator operation. Preferably, the server 200 is configured to assign passengers with the same or similar destination floors to the same elevator based on a neural network algorithm, with the constraint that the waiting time for the reserved persons and elevators is minimized. It should be noted that the elevator scheduling/control strategy obtained by the neural network algorithm is not the optimal scheduling/control strategy, and because the number of high-rise elevators is large, the carried passenger flow rate is large, the entering and exiting conditions in different time periods are different, the scheduling/control strategy on one side of the server 200 needs to be continuously and dynamically changed, and can be fed back in real time along with the change of the elevators and the passenger flow, and the middle influence factors are more, the server 200 is configured to obtain the elevator scheduling/control strategy based on the deep reinforcement learning/reinforcement learning algorithm.

Preferably, the basic process of reinforcement learning is a Markov decision process. The Markov decision process may form a quadruple representation { s, a, p, r } with state s, action a, state transition probability p, state transition reward or reward r. For the discrete-time markov decision process, the set of states and actions is referred to as the state space S and the action space a. Is specifically represented as state s _i ∈S，a _i E.g. A. According to the action selected in step t, the state is according to the probability P(s) _t+1 ,s _t ,a _t ) From s _t Is transferred to s _t+1 . At the same time of state transition, the decision body gets 1 instant reward R(s) _t+1 ,s _t ,a _t ). S in the above expression _t Shown as the state at time t. a is _t The operation at time t is shown. The accumulated rewards at the end of the above process are:

G _t ＝R _t +γR _t+1 +γ ² R _t+2 +…+γ ^k R _t+k ＝∑ _k＝0 γ ^k R _t+k (1)

r in the formula (1) _t Is the prize accumulated over time t. Gamma is a discount factor, and the value range is between 0 and 1. The discounting factor is used to reduce the reward weight corresponding to the future decision. The ultimate goal of the decision is to achieve maximization of the jackpot while reaching the goal state.

Preferably, the state space of the embodiment is constructed by the state of the elevator operation collected by the internet of things device 100, the position and arrival time of the reservation personnel, and the like. It should be noted that the internet of things device 100 can waitAnd monitoring by reservation personnel. The dispatch/control of elevators by the server 200 is the action space. The optimization goals of the strategy for elevator dispatching/control of the embodiment are that the waiting time is shortest and the energy consumption is lowest. Preferably, attributes relating to the shortest waiting times of the elevators and the reservation staff and the energy consumption of the elevators can be selected as the state space. The action of the server 200 controlling the elevator can be taken as a policy variable. Preferably, the number of stops of the elevator and the distance traveled by the elevator can be selected as the movement space. Preferably, the reinforcement learning update strategy is based on a cost function. Preferably, the cost function can be evaluated by constraints. The variable of the cost function is a parameter s in the state space _t And a parameter a in the motion space _t . Preferably, the constraint condition may be that elevator operation efficiency is not reduced and the plurality of elevator car load differences are less than the third threshold. Preferably, the third threshold value can be set according to the number of elevators, the load bearing specification of the elevators, etc. The third threshold value may be 20% in the present embodiment.

Preferably, in the deep reinforcement learning/reinforcement learning, the update function is as follows:

Q(s _t+1 ,a _t+1 )＝Q _o (s _t ,a _t )+loss (2)

q(s) in the formula (2) _t+1 ,a _t+1 ) Is the value of the updated cost function. Q _o (s _t ,a _t ) Previous value in previous state. The previous value is the value stored in the value table. loss is a loss function.

loss＝α[Q _r (s _t+1 ,a _t+1 )-Q _o (s _t ,a _t )] (3)

In formula (3) Q _r (s _t+1 ,a _t+1 ) Is a practical value. α is the learning rate. Alpha is between 0 and 1. Alpha determines the rate of value table update.

Preferably, the real-world value comprises a first real-world value under a first constraint and a second real-world value under a second constraint. The first constraint condition is that the operation efficiency of the elevator is not reduced and the load difference of the elevator cars is smaller than a third threshold value. Preferably, the second constraint is that the waiting time for the booking person, characterized by the first/second positioning message, is minimal.

Preferably, the server 200 configures the first realistic value function as follows:

firstly, recording corresponding cost functions under different states and actions based on a first constraint condition and a second constraint condition respectively, and recording and forming a value table.

And determining the first reality value based on the mode that the value function corresponding to the first constraint condition in the maximum evaluation new state is maximum. Preferably, the first realistic cost function is:

Q _r1 (s _t ,a _t )＝R(s _t ,a _t )+γmaxQ _o (s _t+1 ,a _t+1 ) (4)

q in formula (4) _r1 (s _t ,a _t ) Is a first realistic cost function. R(s) _t ,a _t ) The reward corresponding to the execution of the first action by the execution module 300. Preferably, the reward can be specifically set according to the corresponding policy variable, and generally takes a one-hot form, such as 0, -1, -2, and the like. MaxQ _o (s _t+1 ,a _t+1 ) The maximum value corresponding to the new state in the value table for the new state after the action is executed. Gamma denotes the state s _t By action a _t The value of (c) and the degree of attenuation associated with the sum of the next state and the action. The value range of gamma is between 0 and 1. Preferably, the value table is a discrete record of the cost function.

Preferably, the server 200 configures the second realistic cost function as follows:

preferably, the second realistic value is determined based on the value of the cost function under the second constraint at the new environmental state in the value table. Preferably, the second realistic cost function is:

Q _r2 (s _t ,a _t )＝R(s _t ,a _t )+γQ _o (s _t+1 ,a _t+1 ) (5)

q in formula (5) _r2 (s _t ,a _t ) Representing a second realistic cost function. Q _o (s _t+1 ,a _t+1 ) The corresponding value of the new state in the value table under the second constraint condition after the action is executed is represented. Preferably, the corresponding value may not be

Preferably, the loss function is a linear superposition of the first realistic value and the second realistic value. By adding the first positioning message/the second positioning message of the reservation personnel in the learning update, uncontrollable factors such as the arrival time, the waiting time and the like of the reservation personnel can be considered in the current scheduling/control strategy of the elevator, so that the currently formulated scheduling/control strategy can be dynamically corrected in real time to resist the unexpected operation of the elevator.

Example 2

This embodiment is a supplement and/or improvement to embodiment 1, and repeated contents are not described again.

Preferably, the first positioning message in embodiment 1 may be generated by the internet of things device 100 capable of indoor positioning. Preferably, the internet of things device 100 capable of indoor positioning may be a radio frequency communication device and an optical communication device. The radio frequency communication equipment can realize positioning by utilizing Wi-Fi, bluetooth, infrared, ultrasonic, ultra wide band and RFID technologies. Preferably, in this embodiment, the radio frequency communication device may use Received Signal Strength (RSS) for positioning. Preferably, the radio frequency communication device can also be positioned by means of location fingerprints. The electronic equipment of the reservation personnel is connected with the radio frequency communication equipment. The signal characteristic of the electronic device connected to the radio frequency communication device may be used as the first positioning message. The signal characteristics of the electronic device connected to the radio frequency device include received signal strength, multipath structure of the signal, round trip time and time delay, etc. Preferably, the optical communication device can also perform indoor positioning based on received signal strength, multipath structure of the signal, round trip time and time delay, etc. Typically, an optical communication device may construct an optical communication link with a light sensing element of an electronic device. The light sensing elements may be cameras and light sensors of the electronic device. Specifically, the optical communication device actively emits an optical signal, and the electronic device passively receives an optical signal. Preferably, the optical communication device can also be positioned by imaging through an image sensor. However, the radio frequency communication apparatus performs positioning using the strength of the received signal, and the positioning accuracy is poor and the strength fluctuation of the received signal is large. The radio frequency communication equipment is positioned by adopting a position fingerprint mode, a large amount of position fingerprint information needs to be collected in advance to construct a position fingerprint database, when a user is in a corresponding area, a mobile phone can be used for capturing signal characteristics and comparing the signal characteristics with the database to obtain positioning, and the positioning precision is usually very high. But the workload at the initial stage of database construction is large, and once the environment such as Wi-Fi base stations changes the original database, the database will fail. The communication distance of the optical communication equipment for indoor positioning is limited, generally within 5 meters, and the error rate is high, so that the positioning effect cannot be ensured.

The embodiment provides an intelligent elevator management system based on an intelligent gateway, which comprises an internet of things device 100, a server 200 and an intelligent gateway 300. The intelligent gateway 300 is configured to:

requesting the internet of things device 100 to feed back second information and first positioning messages about an action path generated based on the reserved first information in a manner of periodically accessing the plurality of internet of things devices 100 in a remote configuration of the server 200 and/or the internet of things device 100, and transmitting the second information and the first positioning messages to the server 200;

in the process of periodic access, at least two pieces of internet-of-things equipment 100 which generate the first positioning message and have a signal strength difference exceeding a first threshold value are searched in a one-by-one comparison mode on the basis of the internet-of-things equipment 100 which sends the second information. Preferably, the first threshold may be set according to the installation location of the internet of things device and the actually measured strength of the radio frequency signal. Typically, the first threshold is at least 20%. By adopting the arrangement mode, the invention achieves the beneficial effects that:

the invention can combine the long-time off-line second information and the short-time real-time first positioning information for positioning. This approach can overcome the drawbacks of the prior art positioning techniques. The prior art can use radio frequency communication equipment or optical communication equipment for positioning, but has the following two problems:

1. the radio frequency communication equipment is used for positioning, the received signal strength is generally adopted for positioning, the radio frequency signal of the positioning method is easily influenced by multi-channel fading, and the positioning accuracy is poor and generally exceeds more than ten meters.

2. By adopting the position fingerprint method, a large number of anchor points need to be constructed in the early stage, and the positions of related equipment need to be kept unchanged, so that the later maintenance construction is not facilitated.

Therefore, the invention searches for at least two pieces of internet-of-things equipment 100 with larger positioning difference in the first positioning message in a one-by-one comparison mode by traversing all pieces of internet-of-things equipment 100 through the intelligent gateway 300, and can greatly improve the positioning precision and ensure the continuity of positioning. Specifically, the first positioning message includes location information and signal strength information of the reservation person. The invention firstly analyzes the signal intensity, can obtain the position information for analysis without decoding the consumed computing resource, and can reduce the computing overhead and the time cost at the side of the server 200. On the one hand, the difference of the signal strengths, that is, when the difference of the two signal strengths exceeds the first threshold, proves that the difference of the positioning angles of the two internet of things devices 100 with respect to the reservation personnel is large enough, so that the reservation personnel can be positioned from different angles as much as possible. The method for positioning the reservation personnel from different angles ensures that the positioning result of the reservation personnel has stronger diversity and can improve the reliability for subsequent positioning. For example, when the signal strength of one of the internet of things devices 100 is unstable and has irregular jump, the finally obtained location information is erratic, and if the mean value is taken and then the location is performed, the real-time effect of location is poor. According to the invention, the signal intensity of another piece of Internet of things equipment 100 about the first positioning message is used as a reference value of the jump signal, the first threshold value is used as a correction range, the jump signal beyond the first threshold value range of the reference value is eliminated, and then the eliminated signal is processed by combining Kalman filtering. The processing mode can improve the reliability of the positioning signal on the basis of ensuring the real-time effect.

On the other hand, when the signal strength is stable, the overlapping positioning range of the two first positioning messages with different signal strengths is used as the position of the reservation personnel, so that the positioning accuracy can be greatly improved. Moreover, the positioning accuracy of a single internet of things device 100 is poor, and the device is not sensitive to the movement of reservation personnel. Since the positions of the two internet of things devices 100 are known, when the difference between the signal strengths of the two internet of things devices 100 continuously and stably changes, the movement of the reservation person can be quickly sensed.

In addition, because the position of the internet of things device 100 is fixed, the communication distance for positioning is limited, and the action path of the booked person is likely to be separated from the communication distance of the internet of things device 100, the invention adopts a mode of searching for the signal intensity difference exceeding the first threshold value one by one to perform positioning, and on one hand, can ensure that the continuity of the first positioning messages with two different signal intensities is maintained during the positioning of the booked person, thereby reducing the positioning error and ensuring the real-time performance of positioning. On the other hand, the actual physical position information of different Internet of things equipment can be utilized to provide positioning results at different angles to improve reliability.

Preferably, the intelligent gateway 300 is configured to merge the second information and the first positioning message according to the following steps:

based on the internet of things devices 100 sending the second information, at least two internet of things devices 100 which generate the first positioning messages and have the signal strength difference of the first positioning messages exceeding a first threshold value are searched in a one-by-one comparison mode. Preferably, starting with the physical location of the internet of things device 100 that first sends the second information, the internet of things devices 100 are requested to feed back the first positioning message one by one according to the pre-stored IDs of the internet of things devices 100 sorted by the actual physical location. In the process of searching for at least two pieces of internet of things equipment 100 which generate the first positioning messages and the signal difference of the first positioning messages exceeds a first threshold value, the internet of things equipment 100 is requested to feed back the second information periodically. And if the second information is acquired, adjusting the currently acquired first positioning message based on the second information to avoid the position information transmitted by the first positioning message from deviating from the actual position.

If the difference of the signal strength of transmission is too big, then it is meaningless to reduce the positioning accuracy error, and the signal strength is too strong moreover, and its undulant scope is bigger to when leading to subsequent according to another positioning signal as the benchmark to carry out Kalman filtering, lead to most signals not in first threshold value scope, thereby make most signals filtered.

Preferably, in the case that the difference between the signal strengths of at least two first positioning messages exceeds the second threshold value in the plurality of internet of things devices 100 transmitting the first positioning messages, the intelligent gateway 300 is configured to:

shielding the internet of things equipment 100 transmitting the first positioning message, wherein the signal intensity of the internet of things equipment 100 is smaller than a fourth threshold value, and requesting to close the internet of things equipment 100;

the internet of things device 100 which does not transmit the first positioning message is searched by using the ID of the internet of things device 100 as an initial point, so that the intelligent gateway 300 keeps communication with at least two internet of things devices 100 which transmit the first positioning message and have the signal strength difference of the first positioning message exceeding a first threshold value. Preferably, the second threshold is 50%. The fourth threshold may be set according to the actual location and installation parameters of the internet of things device 100. Preferably, the fourth threshold is typically-70 dBm to-85 dBm.

Preferably, in the case that the difference between the signal strengths of at least two first positioning messages exceeds the second threshold value and there is no internet of things device 100 transmitting the first positioning message with a signal strength less than the fourth threshold value among the plurality of internet of things devices 100 transmitting the first positioning message, the smart gateway 300 is configured to:

shielding the internet of things equipment 100 which transmits the first positioning message and has the signal strength greater than a fifth threshold, and requesting to close the internet of things equipment 100;

the internet of things device 100 which does not transmit the first positioning message is searched by using the ID of the internet of things device 100 as an initial point, so that the intelligent gateway 300 keeps communication with at least two internet of things devices 100 which transmit the first positioning message and have the signal strength difference of the first positioning message exceeding a first threshold value. Preferably, the fifth threshold may be set according to the indoor structure and the specific location of the internet of things device 100. Generally, the fifth threshold may be above-45 dBm if a Wi-Fi signal is used for positioning. Through this setting mode, the beneficial effect who reaches is:

the strength of the first positioning message signal fed back by the internet of things device 100 is kept between the fourth threshold and the fifth threshold, and the difference between the strengths of at least two first positioning message signals is kept between the first threshold and the second threshold, so that most of signals are prevented from being filtered due to too large signal fluctuation.

Example 3

The invention also provides an intelligent elevator management method based on the intelligent gateway. As shown in fig. 2, the method comprises the steps of:

s100: and transmitting the first information acquired based on the reservation information to the internet of things device 100 through the intelligent gateway 300. Preferably, the first information may include a destination floor of the reservation person, physiological information of the reservation person, authority information of the reservation person, identification information of an electronic device of the reservation person, and the like. The physiological information of the reservation personnel comprises face information, fingerprint information and the like.

S200: requesting, in a remote configuration of the server 200 and/or the internet of things devices 100, the internet of things devices 100 to feed back second information and first location messages about an action path generated based on the reserved first information in a manner of periodically accessing the plurality of internet of things devices 100, and transmitting the second information and the first location messages to the server 200. In the process of periodic access, at least two pieces of internet-of-things equipment 100 which generate the first positioning message and have a signal strength difference exceeding a first threshold value are searched in a one-by-one comparison mode on the basis of the internet-of-things equipment 100 which sends the second information. Preferably, the first threshold may be set according to the installation location of the internet of things device and the actually measured strength of the radio frequency signal. Typically, the first threshold is at least 20%. By adopting the arrangement mode, the invention achieves the beneficial effects that:

according to the invention, all the Internet of things devices 100 are traversed through the intelligent gateway 300, and at least two Internet of things devices 100 with larger positioning difference in the first positioning message are searched in a one-by-one comparison mode, so that the positioning precision can be greatly improved and the positioning continuity can be ensured. Specifically, the first positioning message includes location information and signal strength information of the reservation person. The invention firstly analyzes the signal strength, can obtain the position information for analysis without decoding the consumed computing resource, and can reduce the computing overhead and time cost on one side of the server 200. On the one hand, the difference of the signal strengths, that is, when the difference of the two signal strengths exceeds the first threshold, proves that the difference of the positioning angles of the two internet of things devices 100 with respect to the reservation personnel is large enough, so that the reservation personnel can be positioned from different angles as much as possible. The method for positioning the reservation personnel from different angles ensures that the positioning result of the reservation personnel has stronger diversity and can improve the reliability for subsequent positioning. For example, when the signal strength of one of the internet of things devices 100 is unstable and has irregular jump, the finally obtained location information is erratic, and if the mean value is taken and then the location is performed, the real-time effect of location is poor. According to the invention, the signal intensity of another piece of Internet of things equipment 100 about the first positioning message is used as a reference value of the jump signal, the first threshold value is used as a correction range, the jump signal beyond the first threshold value range of the reference value is eliminated, and then the eliminated signal is processed by combining Kalman filtering. The processing mode can improve the reliability of the positioning signal on the basis of ensuring the real-time effect.

On the other hand, when the signal strength is stable, the positioning accuracy can be greatly improved by using the overlapping positioning range of the two first positioning messages with different signal strengths as the position of the reservation person. Moreover, the positioning accuracy of a single internet of things device 100 is poor, and the device is not sensitive to the movement of reservation personnel. Since the positions of the two internet of things devices 100 are known, when the difference between the signal strengths of the two internet of things devices 100 continuously and stably changes, the movement of the reservation person can be quickly sensed.

In addition, because the position of the internet of things device 100 is fixed, the communication distance for positioning is limited, and the action path of the reservation person is likely to be separated from the communication distance of the internet of things device 100, the invention performs positioning by searching for the signal intensity difference one by one to exceed the first threshold, on one hand, the continuity of the first positioning messages with two different signal intensities can be ensured to be kept during the positioning of the reservation person, thereby reducing the positioning error and ensuring the positioning real-time performance. On the other hand, the actual physical position information of different Internet of things equipment can be utilized to provide positioning results at different angles to improve reliability.

If the difference in the signal strength of the transmissions is too large, it is not intended to reduce the positioning accuracy error, and the signal strength is too strong, and the fluctuation range is larger, so that when the kalman filtering is performed based on another positioning signal, most of the signals are not within the first threshold range, and most of the signals are filtered.

Preferably, in the case that, among the plurality of internet of things devices 100 transmitting the first positioning message, the difference between the signal strengths of at least two first positioning messages exceeds the second threshold value and there is no internet of things device 100 transmitting the first positioning message with a signal strength smaller than the fourth threshold value, the smart gateway 300 is configured to:

S300: planning a navigation path for the reservation personnel and dynamically adjusting a dispatching/control strategy of the elevator at least based on second information and first positioning information which are uploaded by the Internet of things equipment 100 and related to the action path and are triggered by the first information. And/or navigating the reservation personnel and dynamically adjusting the dispatching/control strategy of the elevator based on the second positioning information uploaded by the reservation personnel. Preferably, the second information includes at least ID information and time information of the internet of things device 100. The time information is time information generated by the internet of things device 100 based on the triggering of the first information. Preferably, the server 200 may obtain the location information through ID information of the internet of things device 100. Since the location of the internet of things device 100 is fixed, the location information of the internet of things device 100 can be mapped one to one by compiling an ID in advance. For example, when the internet-of-things device 100 is an access control device, the access control device verifies the identity of a reservation person based on the first information transmitted from the server 200. The entrance guard equipment can verify the identity of the reservation personnel through face recognition, fingerprint recognition and electronic equipment of the reservation personnel. The identity verification by using the electronic equipment of the reservation personnel can be that the server 200 sends the two-dimensional code corresponding to the first information, when the reservation personnel pass through the access control equipment, the access control equipment scans the two-dimensional code displayed by the electronic equipment of the reservation personnel and then decrypts, and if the decrypted information is consistent with the first information, the identity of the reservation personnel is judged. Preferably, the first information may be a public key. The access control equipment can utilize the first information to decrypt the two-dimensional code displayed by the reservation personnel. Preferably, the server 200 is configured to anticipate the expected arrival time of the reservation person and the selected elevator based at least on said second information and second positioning message. The server 200 is configured to dynamically adjust the dispatch/control strategy of the elevator based on the expected arrival time and the selected elevator. Preferably, the second positioning message may be a manner in which a reservation person may connect to the server 200 through the internet of things device 100 or an intelligent gateway, and the like, and actively upload the location information thereof. Through the above setting mode, the beneficial effect who reaches is:

the prior art does not consider the relation between an elevator and position navigation, on one hand, the position setting or elevator planning of the elevator is not standard, and a reservation person may not find the elevator or the selected elevator cannot reach the target floor, so the position navigation function is needed. On the other hand, the intelligent dispatching and control of the elevator not only needs to monitor the running state of the elevator, but also needs to reserve the exact arrival time of the personnel for planning so as to reduce the operation of unexpected plans. That is, the running state of the elevator itself is changed from moment to moment, and if the time exactly reached by the reservation personnel is not available, the currently established or expected elevator service path is not in accordance with the reality, thereby causing the unexpected running of the elevator, increasing the abrasion of equipment and the consumption of energy. The elevator management system of the present invention uses the internet of things of the elevator, i.e., the server 200 acquires the location information and time information of the reservation person in non-real time and offline by sensing the fixed internet of things devices 100. For example, when a reservation person passes through the access control device, the access control device allows the reservation person to pass through based on first information of the reservation person, and feeds back the time when the reservation person passes through the access control device and the ID of the reservation person to the server 200, so that the server 200 can know the action track of the reservation person. The server 200 can quickly plan the navigation path of the reservation staff and guide the reservation staff to reach the designated elevator. Meanwhile, the action track of the reservation personnel acquired through the internet of things equipment 100 similar to the access control equipment is non-real-time offline position information, on one hand, the navigation path planned by the server 200 may deviate from the actual action path of the reservation personnel due to the hysteresis of acquiring the position information of the reservation personnel, on the other hand, the server 200 cannot correct the deviation of the reservation personnel when the reservation personnel deviates from the navigation path or does not select a corresponding elevator because the position information of the reservation personnel cannot be acquired in real time, and on the other hand, the planned elevator dispatching scheme or control scheme cannot be adjusted and corrected in real time, so that a large amount of unexpected operation of the elevator is caused. Therefore, the invention acquires the first positioning message or the second positioning message by utilizing the indoor positioning function in the internet of things device 100 to acquire the real-time online position information of the reservation personnel, and corrects the real-time online position information by the offline position information fixed by the internet of things, so as to acquire the expected arrival time of the reservation personnel and navigate for the reservation personnel to reduce the uncertainty of selecting the elevator for the reservation personnel, and further the server 200 can integrate the arrival time of the reservation personnel and the certainty of selecting the elevator when determining the scheduling/control strategy of the elevator, so that the generated scheduling/control strategy can reduce the unexpected operation of the elevator, and further avoid the increase of energy consumption and the increase of equipment wear.

acquiring first information about a reservation;

a dispatch/control strategy for the elevator is anticipated based on the first information and the elevator traffic demand. Preferably, the first information comprises a destination address provided by the reservation operator. The destination address includes a destination floor to which the reservation person is to reach. Preferably, the server 200 may employ a fuzzy algorithm for logic processing to sense peak elevator traffic. The server 200 may also be configured to obtain the demand for elevator traffic based on a risk prediction of elevator operation. Preferably, the server 200 is configured to assign passengers with the same or similar destination floors to the same elevator based on a neural network algorithm, with the constraint that the waiting time for the reserved persons and elevators is minimized. It should be noted that the elevator scheduling/control strategy obtained by the neural network algorithm is not an optimal scheduling/control strategy, and because there are many elevators in a high-rise building, the amount of people carried by the elevators is large, and the entering and exiting conditions in different time periods are different, the scheduling/control strategy on one side of the server 200 needs to be continuously and dynamically changed, and can be fed back in real time along with the changes of the elevators and the people flows, and there are many middle influence factors, the server 200 is configured to obtain the elevator scheduling/control strategy based on the deep reinforcement learning/reinforcement learning algorithm.

Preferably, the basic process of reinforcement learning is a markov decision process. The Markov decision process may form a quadruple representation { s, a, p, r } with state s, action a, state transition probability p, state transition reward or reward r. For the discrete-time markov decision process, the set of states and actions is referred to as the state space S and the action space a. Is expressed in detail as state s _i ∈S，a _i E.g. A. According to the action selected in step t, the state is according to the probability P(s) _t+1 ,s _t ,a _t ) From s _t Is transferred to s _t+1 . At the same time of state transition, the decision body receives 1 instant rewardR(s _t+1 ,s _t ,a _t ). S in the above expression _t Shown as the state at time t. a is _t Shown as the action at time t. The prizes accrued at the end of the above process are:

r in the formula (1) _t Is the accumulated prize in time t. Gamma is a discount factor, and the value range is between 0 and 1. The discounting factor is used to reduce the reward weight corresponding to the future decision. The ultimate goal of the decision is to achieve maximization of the jackpot while reaching the goal state.

Preferably, the state space of the embodiment is constructed by the state of the elevator operation collected by the internet of things device 100, the position and arrival time of the reservation personnel, and the like. It should be noted that the internet of things device 100 may monitor the waiting reservation staff. The dispatch/control of elevators by the server 200 is the action space. The optimization goals of the strategy for elevator dispatching/control of the embodiment are that the waiting time is shortest and the energy consumption is lowest. Preferably, attributes relating to the shortest waiting times of the elevators and the reservation staff and the energy consumption of the elevators can be selected as the state space. The action of the server 200 controlling the elevator can be taken as a policy variable. Preferably, the number of stops of the elevator and the distance traveled by the elevator can be selected as the movement space. Preferably, the reinforcement learning update strategy is based on a cost function. Preferably, the cost function can be evaluated by constraints. The variables of the cost function being parameters s in the state space _t And a parameter a in the motion space _t . Preferably, the constraint condition may be that elevator operation efficiency is not reduced and the plurality of elevator car load differences are less than the third threshold. Preferably, the third threshold value can be set according to the number of elevators, the load bearing specification of the elevators, etc. The third threshold value may be 20% in the present embodiment.

Q(s _t+1 ,a _t+1 )＝Q _o (s _t ,a _t )+loss (2)

loss＝α[Q _r (s _t+1 ,a _t+1 )-Q _o (s _t ,a _t )] (3)

Q in formula (3) _r (s _t+1 ,a _t+1 ) Is a practical value. α is the learning rate. Alpha is between 0 and 1. Alpha determines the rate of value table update.

Preferably, the real-world value comprises a first real-world value under a first constraint and a second real-world value under a second constraint. The first constraint condition is that the operation efficiency of the elevator is not reduced and the load difference of the elevator cars is less than a third threshold value. Preferably, the second constraint is that the waiting time for the booking person, characterized by the first/second positioning message, is minimal.

Preferably, the server 200 configures the first realistic cost function as follows:

And determining the first reality value based on the mode that the maximum evaluation value of the cost function corresponding to the first constraint condition in the new state is the maximum. Preferably, the first realistic cost function is:

Q _r1 (s _t ,a _t )＝R(s _t ,a _t )+γmaxQ _o (s _t+1 ,a _t+1 ) (4)

preferably, the second realistic value is determined based on the value of the cost function under the second constraint under the new environmental state in the value table. Preferably, the second realistic cost function is:

Q _r2 (s _t ,a _t )＝R(s _t ,a _t )+γQ _o (s _t+1 ,a _t+1 ) (5)

in formula (5) Q _r2 (s _t ,a _t ) Representing a second realistic cost function. Q _o (s _t+1 ,a _t+1 ) The corresponding value of the new state in the value table under the second constraint condition after the action is executed is represented. Preferably, the corresponding value may not be

Preferably, the loss function is a linear superposition of the first and second real-world worth values. By adding the first positioning message/the second positioning message of the reservation personnel in the learning update, uncontrollable factors such as the arrival time, the waiting time and the like of the reservation personnel can be considered in the current scheduling/control strategy of the elevator, so that the currently formulated scheduling/control strategy can be dynamically corrected in real time to resist the unexpected operation of the elevator.

The present specification encompasses multiple inventive concepts and the applicant reserves the right to submit divisional applications according to each inventive concept. The present description contains a plurality of inventive concepts such as "preferably", "according to a preferred embodiment" or "optionally" each indicating that the respective paragraph discloses a separate concept, the applicant reserves the right to apply for divisional applications according to each inventive concept.

It should be noted that the above-mentioned embodiments are exemplary, and that those skilled in the art, having benefit of the present disclosure, may devise various arrangements that are within the scope of the present disclosure and that fall within the scope of the invention. It should be understood by those skilled in the art that the present specification and figures are illustrative only and are not limiting upon the claims. The scope of the invention is defined by the claims and their equivalents.

Claims

1. An intelligent elevator management system based on an intelligent gateway is characterized by comprising an Internet of things device (100), a server (200) and the intelligent gateway (300), wherein,

the server (200) is configured to:

transmitting first information related to reservation personnel acquired based on reservation information to the Internet of things device (100) through the intelligent gateway (300);

planning a navigation path for a reservation person and dynamically adjusting a dispatching/control strategy of an elevator at least based on second information which is uploaded by the Internet of things device (100) and records ID information and time information of the Internet of things device (100) and first positioning information which records position information and signal strength information of the reservation person, wherein the second information is related to an action path and records ID information and time information of the Internet of things device (100) and the first positioning information records position information and signal strength information of the reservation person; and/or

Navigating the reservation personnel and dynamically adjusting the dispatching/control strategy of the elevator based on a second positioning message which is uploaded by the reservation personnel and records the position information of the reservation personnel;

the intelligent gateway (300) is configured to:

requesting, at a remote configuration of the server (200) and/or the internet of things device (100), the internet of things device (100) to feed back second information describing ID information and time information of the internet of things device (100) and a first positioning message describing location information and signal strength information of a reservation person on an action path generated based on reserved first information related to the reservation person, and transmitting the second information and the first positioning message to the server (200), in a manner of periodically accessing a plurality of the internet of things devices (100);

in the process of periodic access, searching at least two pieces of Internet of things equipment (100) which generate the first positioning messages and have the signal strength difference exceeding a first threshold value with each other in a comparison mode one by one on the basis of the Internet of things equipment (100) which sends out the second information;

the Internet of things device (100) is configured to:

feeding back second information on offline path of the reservation person to the intelligent gateway (300) based on the first information on the reservation transmitted by the intelligent gateway (300), wherein,

and issuing an RF signal and/or an optical signal under the instruction of the intelligent gateway (300) and/or the triggering of the second information so as to acquire the RF signal and/or the optical signal which are fed back by the electronic equipment of the reservation personnel and are related to the online position, and transmitting the fed-back RF signal and/or the optical signal to the intelligent gateway (300).

2. The intelligent elevator management system according to claim 1, characterized in that the server (200) is configured to anticipate the dispatch/control strategy of an elevator as follows:

acquiring first information about a reservation;

predicting elevator traffic demand based on real-time elevator running states and historical elevator running states fed back by the Internet of things equipment (100);

3. The intelligent elevator management system according to claim 1, characterized in that the server (200) is configured to obtain the dispatch/control strategy of an elevator based on a deep reinforcement learning/reinforcement learning algorithm, wherein,

the optimization target is as follows: the waiting time is shortest and the energy consumption is lowest;

the constraint conditions are as follows: the elevator operation efficiency is not reduced, and the load difference of the elevator cars is smaller than a third threshold value;

the strategy variables are: the stopping times and the running distance of the elevator;

wherein, in the deep reinforcement learning/reinforcement learning update function, a loss function with respect to latency characterized by the first positioning message/the second positioning message is added.

4. The intelligent elevator management system according to claim 1, wherein the intelligent gateway (300) is configured to fuse the second information and the first positioning message according to the following steps:

searching, on the basis of the IOT device (100) issuing the second information, at least two IOT devices (100) which generate the first positioning message and whose signal strength differences of the first positioning messages exceed a first threshold value in a comparison-by-comparison manner, wherein,

starting with the physical position of the Internet of things equipment (100) which firstly sends out the second information, requesting the Internet of things equipment (100) to feed back the first positioning message one by one according to the pre-stored IDs of the Internet of things equipment (100) which are sorted by the actual physical position;

periodically requesting the IOT device (100) to feed back second information in the process of searching at least two IOT devices (100) which generate the first positioning message and have the signal difference of the first positioning message exceeding a first threshold value, wherein,

and if the second information is acquired, adjusting the currently acquired first positioning message based on the second information to avoid the position information transmitted by the first positioning message from deviating from the actual position.

5. The intelligent elevator management system according to claim 1, wherein in case of a signal strength difference of at least two first positioning messages exceeding a second threshold value in a plurality of internet of things devices (100) transmitting the first positioning messages, the intelligent gateway (300) is configured to:

shielding the Internet of things equipment (100) transmitting the first positioning message, wherein the signal intensity of the Internet of things equipment (100) is smaller than a fourth threshold value, and requesting to close the Internet of things equipment (100);

and searching the IOT equipment (100) which does not transmit the first positioning message by taking the ID of the IOT equipment (100) as the start so that the intelligent gateway (300) is kept in communication with at least two IOT equipment (100) which transmit the first positioning message and the signal strength difference of the first positioning messages exceeds a first threshold value.

6. The intelligent elevator management system according to claim 1, wherein in case that there is at least two first positioning messages transmitted by the internet of things devices (100) having signal strength difference exceeding the second threshold and there is no internet of things device (100) having signal strength of the first positioning message less than the fourth threshold, the intelligent gateway (300) is configured to:

shielding the Internet of things equipment (100) transmitting the first positioning message, wherein the signal intensity of the Internet of things equipment is greater than a fifth threshold value, and requesting to close the Internet of things equipment (100);

7. An intelligent elevator management method based on an intelligent gateway is characterized by comprising the following steps:

transmitting first information related to reservation personnel acquired based on reservation information to the Internet of things equipment (100) through the intelligent gateway (300);

Navigating the reservation personnel and dynamically adjusting the dispatching/control strategy of the elevator based on the second positioning message which is uploaded by the reservation personnel and records the position information of the reservation personnel;

wherein the smart gateway (300) requests the IOT device (100) to feed back second information and first positioning messages about an action path generated based on the reserved first information in a manner of periodically accessing a plurality of IOT devices (100) under the remote configuration of the server (200) and/or the IOT devices (100), and transmits the second information and the first positioning messages to the server (200);

in the process of periodic access, at least two Internet of things devices (100) which generate the first positioning messages and have the signal strength difference exceeding a first threshold value are searched in a comparison-by-comparison mode on the basis of the Internet of things devices (100) which send out the second information.

8. The intelligent elevator management method according to claim 7, characterized in that the server (200) is configured to anticipate the dispatch/control strategy of an elevator as follows:

acquiring first information about a reservation;