CN107463154B - Intelligent hotel energy-saving control method and system - Google Patents

Abstract

The present invention provides an energy-saving control method and system for an intelligent hotel. The method is applied to an energy-saving control server, including: acquiring multiple pieces of historical check-in information of a customer, generating a comfort level model corresponding to the customer based on all the obtained historical check-in information, and returning Time probability model; the check-in information includes: the room occupancy information of the customer; the room occupancy information of the customer includes: room occupancy status, room entry and exit time, and energy-saving equipment status information; when the room occupancy status of the customer's room changes, Use the comfort level model and return time probability model corresponding to the customer to adjust the state of the energy-saving equipment in the customer's room, so as to realize energy-saving control on the premise of ensuring customer comfort.

Description

A kind of intelligent hotel energy-saving control method and system

technical field

The invention relates to the technical field of intelligent hotels, in particular to an energy-saving control method and system for an intelligent hotel.

Background technique

Smart Hotel is the intelligence of the hotel. Hotel intelligence is an information network system with new communication technology, computer intelligent information processing, and broadband interactive multimedia network technology as the core. Relying on the intelligent hotel management system, it integrates intelligent lighting management, air conditioning management, call management and information service management functions. Integrated, through the Internet of Things technology, the hotel's various software and hardware are better connected, providing guests with information, intelligence, and personalized services, improving hotel management, reducing hotel operating costs, and improving guest satisfaction.

Existing smart hotels install various intelligent energy-saving hardware devices, as well as various wired and wireless sensors. The existing smart hotel management system mainly relies on various wired and wireless sensors in the room to sense environmental information and customer behavior, and use these sensed information as the input of the equipment control module to improve user comfort and save energy. However, this method does not take into account the direct or indirect relationship between customer comfort, energy saving rate and multiple dimensions of data. The input data dimension is relatively single, and the data processing is too simple and direct, resulting in The value of a large amount of data has not been effectively exploited, and the control method of the equipment is relatively simple, and the balance between customer comfort and energy saving is poor.

SUMMARY OF THE INVENTION

In view of this, the purpose of the present invention is to provide a smart hotel energy saving control method and system, which can improve the energy saving rate while ensuring the comfort of customers.

In order to achieve the above object, the present invention provides the following technical solutions:

An energy-saving control method for an intelligent hotel, applied to an energy-saving control server, comprising:

Acquire multiple pieces of historical check-in information of a customer, and generate a comfort level model and a return time probability model corresponding to the customer based on all the obtained historical check-in information; the check-in information includes: the room check-in information of the customer; the room check-in information of the customer includes: Room occupancy status, room entry and exit time, and energy-saving equipment status information;

When the room occupancy state of the customer's room changes, the state of the energy-saving equipment in the customer's room is adjusted by using the comfort level model and the return time probability model corresponding to the customer to realize energy-saving control.

An energy-saving control system for an intelligent hotel, the system includes an energy-saving control server,

The energy-saving control server is used for acquiring multiple pieces of historical check-in information of a customer, and generating a comfort level model and a return time probability model corresponding to the customer based on all the obtained historical check-in information; the check-in information includes: the room check-in information of the customer; The room occupancy information of the customer includes: room occupancy status, room entry and exit time, and energy-saving equipment status information; when the room occupancy status of the customer's room changes, the comfort level model corresponding to the customer and the return time probability model are used to compare the data. The state of the energy-saving equipment in the customer's room is adjusted to achieve energy-saving control.

It can be seen from the above technical solutions that in the present invention, the obtained multiple pieces of customer occupancy information are used to generate the comfort level model and the return time probability model corresponding to the customer, so that when the occupancy status of the customer's room changes, the comfort level corresponding to the customer is used. The degree model and the return time probability model are used to control the energy-saving equipment in the customer's occupancy area, so that the energy consumption can be reduced to the greatest extent and the energy-saving purpose can be achieved under the condition of ensuring the comfort of the customer.

Description of drawings

1 is a schematic diagram of an intelligent hotel network system according to an embodiment of the present invention;

FIG. 2 is a flowchart of an energy-saving control method for a smart hotel according to an embodiment of the present invention.

Detailed ways

In order to make the objectives, technical solutions and advantages of the present invention clearer, the technical solutions of the present invention will be described in detail below with reference to the accompanying drawings and embodiments.

In the present invention, for each customer, the historical check-in information of the customer is collected, and the comfort level model and return time probability model corresponding to the customer are generated according to the historical check-in information. The energy-saving equipment in the customer's current room is controlled for energy-saving, so as to save the energy to the greatest extent on the premise of ensuring the comfort of the customer.

The realization principle of the present invention is described below:

1. Data collection

In the embodiment of the present invention, the check-in information of the customer mainly includes the following contents: customer information, room information of the customer, and external environment information. in,

Customer information mainly includes: gender, age, nationality, check-in time, check-in room, check-in length, personal preferences (such as preferring high temperature, preferring to smoke), etc.

The information about the customer's room includes: room occupancy status, room entry and exit time, and energy-saving equipment status information. Among them, the room occupancy status includes: check in (check in), occupied (occupied), unoccupied (unoccupied), check out (check out); room entry and exit time indicates the time when customers enter the room or leave the room; energy-saving equipment is Refers to the equipment configured in the customer's room, and the energy consumption can be adjusted by adjusting the status of the equipment.

The external environment information includes: weather information, etc. The weather information mainly includes city ID, temperature, humidity, wind speed, weather description information, and the like.

The acquisition of the above information will be described below with reference to FIG. 1 .

Referring to FIG. 1, FIG. 1 is a schematic diagram of a smart hotel network system according to an embodiment of the present invention. The smart hotel network system includes a property management server (Property Management System, PMS) system, an energy-saving control server, and a gateway (configured for each room in the smart hotel). From gateway 1 to gateway N in Figure 1, each gateway corresponds to a room in a smart hotel), and the gateway of each room is also connected to the energy-saving device in the room (the energy-saving device in a room can have one or more Multiple, in Figure 1, one energy-saving device is used to represent all energy-saving devices in the room), sensors (sensors can include door sensor sensor, motion sensor, motion sensor and other sensors, in Figure 1, one sensor is used to represent a room. All sensors configured) are connected through a wireless/wired network, and the sensors are mainly used to sense customer behavior.

In the above-mentioned FIG. 1 , the energy-saving control server and the PMS system may be the same server, or may be different servers. In addition, the smart hotel network system may also include a variety of other devices capable of networking, which are not marked in FIG. 1 because they are not related to the implementation of the present invention.

In the above-mentioned smart hotel network system shown in FIG. 1 , the PMS system is mainly used for the input and reporting of customer information, and the reporting of the occupancy status information of the customer's room.

Reporting of customer information:

In the existing implementation, when a customer checks in, he first needs to check in at the hotel front desk. During the process of customer check-in, the hotel front desk staff enters the customer information in the PMS system and assigns the customer a room number. .

The customer information entered in the PMS system is shown in Table 1:

Table I

After entering the customer information, the PMS system can report the customer information to the energy-saving control server.

Report of the occupancy status information of the customer's room:

After entering the customer information, the PMS system needs to set the occupancy status of the customer's room. When the customer checks out, it also needs to set the occupancy status of the customer's room.

In the embodiment of the present invention, the occupancy status of the room includes check in (check in), occupied (occupied), unoccupied (unoccupied), and check out (check out). in,

Check in indicates that the customer has checked in at the hotel front desk, but has not yet checked into the assigned room;

Occupied indicates that the customer has checked in at the hotel front desk, has also checked into the allocated room, and the customer is currently in the allocated room;

unoccupied indicates that the customer has checked in at the front desk of the hotel and has also checked into the allocated room, and the customer is not currently in the allocated room;

Check out means that the customer has checked out at the hotel front desk.

After entering the customer information in the PMS system, since the customer has not yet checked into the allocated room, the occupancy status of the customer's room is set to check in at this time.

When reporting the entered customer information to the energy-saving control server, the PMS system also reports the occupancy status of the customer's room (check in at this time) to the energy-saving control server at the same time.

After the customer checks out at the hotel front desk, the PMS system needs to set the occupancy status of the customer's room to check out.

After the PMS system sets the occupancy status of the customer's room to check out, the occupancy status of the room needs to be further reported to the energy-saving control server.

In the smart hotel network system shown in Figure 1 above, the gateways corresponding to each room are mainly used for reporting the occupancy status information of the customer's room, the room entry and exit time of the customer's room, and the status information of each energy-saving device in the customer's room.

Report of the occupancy status information of the customer's room:

After the customer has checked in at the hotel front desk, the customer can go in and out of the room assigned to him at any time before the customer checks out.

When the customer opens the room door to enter the room or leave the room, the door sensor configured in the room can sense the state change of the room door, and report the sensed room door state change information to the gateway configured in the room. Here, the room door state change information includes: state change and state change time. Among them, the state changes include: open to closed (that is, the room door changes from the open state to the closed state), closed to open (that is, the room door changes from the closed state to the open state); the state change time is when the door magnetic sensor senses the room door The time of the state change.

When the customer opens the room door to enter the room or leave the room, the motion sensor can also sense the customer's movement information, and report the customer's movement information to the gateway configured in the room. Here, the client movement information includes movement direction, movement speed, movement distance, and the like.

After receiving the room door state change information reported by the door sensor and the customer movement information reported by the motion sensor, the gateway configured in the customer's room can determine the customer behavior information according to the received room door state change information and customer movement information. Here, the customer behavior information includes: entering the room and leaving the room.

In one embodiment of the present invention, the specific method of determining the customer behavior information according to the room door state change information received from the door magnetic sensor and the customer movement information received from the motion sensor by the gateway configured for the customer's check-in room is:

If the state of the room door of the room changes from closed to open, if the moving direction of the customer is to move into the room and the moving distance indicates that the customer has moved into the room, it is determined that the customer behavior information is entering the room, if the moving direction of the customer is If it moves out of the room and the moving distance indicates that the customer has moved out of the room, it is determined that the customer behavior information is to leave the room;

If the state of the room door of the room changes from open to closed, if the customer's moving direction is to move into the room and the moving distance indicates that the customer has moved into the room, it is determined that the customer's behavior information is entering the room, and if the customer's moving direction is Moving out of the room and the moving distance indicates that the customer has moved out of the room, the customer behavior information is determined to be leaving the room.

After the gateway configured in the customer's room determines the customer's behavior information according to the room door state change information received from the door sensor and the customer's movement information received from the motion sensor, it can determine the occupancy status of the customer's room according to the customer's behavior information. When the customer behavior information indicates that the customer enters the room, the occupancy status of the customer's room is determined to be occupied. When the customer behavior information indicates that the customer leaves the room, the occupancy status of the customer's room is determined to be unoccupied.

When the gateway configured in the room where the customer is staying finds that the occupancy status of the room has changed, it can actively report the current occupancy status information of the room to the energy-saving control server.

The report of the room entry and exit time of the customer's room:

In this embodiment of the present invention, the room entry and exit time includes: entry time and departure time.

After the gateway configured for the room where the customer is staying determines the customer behavior information according to the room door state change information received from the door sensor and the customer movement information received from the motion sensor, if the customer enters the room according to the customer behavior information, it can further The entry time of the room is set as the state change time in the room door state change information; if it is determined according to the customer behavior information that the customer leaves the room, then the leave time of the room can be further set as the state change time in the room door state change information. State change time.

When the gateway configured in the room that the customer stays in reports the occupancy status information of the room to the energy-saving control server, it can also report the room entry and exit time of the room to the energy-saving control server at the same time.

Report of the status information of each energy-saving equipment in the customer's room:

Each room of a smart hotel is usually equipped with a variety of energy-saving devices. The energy-saving devices mentioned here refer to devices that can adjust energy consumption by changing device status information, such as thermostats, humidifiers, televisions, refrigerators, etc. Energy consumption can be adjusted by changing the device state. Taking a thermostat as an example, the thermostat state information includes: thermostat setting temperature, thermostat fan mode, cooling setting temperature, and heating setting temperature. The thermostat energy consumption can be reduced or the thermostat energy consumption can be increased by changing the thermostat setting temperature.

In actual use, changes in the status information of energy-saving devices will have a certain impact on the comfort experience of customers. For example, the thermostat setting temperature is too high or too low, which may make customers feel uncomfortable. At this time, it is necessary to adjust the thermostat Set the temperature to adjust, so as to provide customers with a more comfortable temperature environment. Therefore, the change of state information of energy-saving equipment is an important factor affecting comfort.

Therefore, in the present invention, when the state information of any energy-saving device in the customer's room changes, the changed state information of the energy-saving device can be reported to the gateway configured in the room, and then the gateway configured in the room can be It is reported to the energy-saving control server, so that the energy-saving control server uses the energy-saving device state information to generate the comfort level model.

Access to external environmental information:

The energy-saving control server may obtain the weather information of the location of the smart hotel from the application module for providing the weather information. Here, the application program module for providing weather information may be a local function module deployed in the energy-saving control server, or may be a function module deployed on other servers.

The above reports of customer information, room occupancy status information, room entry and exit time, and energy-saving equipment status information are all actively reported when these information change. When the PMS sets the occupancy status of the room where the customer is staying, it actively reports the occupancy status information of the room; when the gateway configured in each room finds that the occupancy status information of the room changes, it actively reports the occupancy status information and room entry and exit time after the change of the room. After the state of the energy-saving equipment in the room changes, it actively reports the changed state information of the energy-saving equipment.

For any room in a smart hotel, when the information actively reported by the PMS system (including customer information, the occupancy status information of the customer's room), and the information actively reported by the gateway configured in the room (including the occupancy status information of the room, Room entry and exit time, energy-saving equipment status information), because the possibility of reporting at the same time is small, so only some information items of customer check-in information are included, and the rest of the information items do not exist (in the present invention, these non-existent information items are referred to as is a vacancy item), so a complete customer check-in information cannot be formed. In order to obtain a complete customer check-in information, usually the most recently obtained customer check-in information or the default value of each information item can be used (in the present invention, it is necessary to pre- Set the legal value range and default value of each information item in the customer check-in information) Complete the information reported by the PMS system or the gateway configured in the room into a complete customer check-in information, for example:

When the energy-saving control server receives the customer information and the occupancy status information of the customer's room reported by the PMS system, in order to complete the information into a piece of user check-in information, since the customer's check-in information has not been obtained at this time, the customer's check-in information can be The default values of the room entry and exit time, energy-saving equipment status information, and external environment information of the check-in room are added to the information reported by the PMS system, thereby forming a complete customer check-in information.

When the energy-saving control server receives the room occupancy status information, room entry and exit time, and energy-saving device status information reported by the gateway of the room where the customer is staying, the server can add the default values of the customer information and external environment information to the room where the customer is staying. In the information reported by the gateway, a complete customer check-in information can be formed; the customer information and external environment information in the customer's check-in information obtained most recently can also be added to the information reported by the gateway of the customer's check-in room, thereby Form a complete customer check-in information.

In practical applications, the energy-saving control server can obtain the complete customer check-in information by means of content complementing when the PMS system and/or the gateway of the customer check-in room actively reports the information, and can also obtain the customer check-in information by means of regular acquisition.

In fact, after the customer checks in at the front desk of the hotel, when the energy saving control server receives the customer information reported by the PMS system, it can start to regularly (for example, every 10 minutes) to obtain the customer's check-in information, and to obtain the customer's check-in information. The occupancy information is stored as the customer's historical occupancy information for subsequent generation of comfort model and return time probability model.

The energy-saving control server can periodically obtain the customer's check-in information from the PMS system, the gateway configured for the customer's check-in room, and the application module for providing weather information. The specific method is: periodically sending a request to the PMS system to obtain the customer's information from the PMS system, Send a request to the network manager of the customer's room configuration to obtain the room's occupancy status information, room entry and exit time and energy-saving device status information from the gateway configured in the customer's room, and obtain the weather of the smart hotel location from the application module for providing weather information Information; aggregate the customer information obtained at the same time, the occupancy status information of the customer's room, the room entry and exit time, the status information of energy-saving equipment, and the weather information of the location of the smart hotel into a piece of customer check-in information.

It should be noted that: after obtaining a piece of customer check-in information, before storing it as the customer's historical check-in information, in addition to filling up the vacancies in the customer check-in information, it is also necessary to determine whether the customer's check-in information contains Whether the value of each information item is within its legal value range, if it is not within the legal value range (for example: the gender of the customer can only be male or female, if it is other values, it is not within the legal value range), Then, the check-in information of the customer can be deleted, so as to avoid introducing erroneous data and causing subsequent generation of a comfort level model and a return time probability model that are not suitable for the customer.

In actual implementation, if the customer has stayed in a smart hotel for many times, the historical occupancy information obtained and stored by the customer when he stayed in the smart hotel can also be used for the comfort model and return time probability model of the customer's current stay in the smart hotel. generate.

The following is an example of historical occupancy information for a stored customer:

A0101,2015/10/01 14:00,1,27,27,Auto,Auto,1,27,19,32,man,N1,8,2015/09/27 20:54,type B,n

A0101,2015/10/01 14:10,1,27,26,Auto,Auto,1,27,19,32,man,N1,8,2015/09/27 20:54,type B,n

A0101,2015/10/01 14:20,1,27,26,Auto,Auto,1,27,19,31,man,N1,8,2015/09/27 20:54,type B,n

A0101,2015/10/01 14:30,1,27,27,Auto,Auto,1,27,19,31,man,N1,8,2015/09/27 20:54,type B,n

A0101,2015/10/01 14:40,1,27,27,Auto,Auto,1,27,19,31,man,N1,8,2015/09/27 20:54,type B,n

A0101,2015/10/01 14:50,1,27,27,Auto,Auto,1,27,19,31,man,N1,8,2015/09/27 20:54,type B,n

A0101,2015/10/01 15:00,1,27,27,Auto,Auto,1,27,19,30,man,N1,8,2015/09/27 20:54,type B,n

A0101,2015/10/01 15:10,1,27,27,Auto,Auto,1,27,19,30,man,N1,8,2015/09/27 20:54,type B,n

A0101,2015/10/01 15:20,1,27,27,Auto,Auto,1,27,19,30,man,N1,8,2015/09/27 20:54,type B,n

A0101,2015/10/01 15:30,1,27,27,Auto,Auto,1,27,19,30,man,N1,8,2015/09/27 20:54,type B,n

A0101,2015/10/01 15:40,1,27,27,Auto,Auto,1,27,19,30,man,N1,8,2015/09/27 20:54,type B,n

A0101,2015/10/01 15:50,1,27,26,Auto,Auto,1,27,19,30,man,N1,8,2015/09/27 20:54,type B,n

A0101,2015/10/01 16:00,1,27,27,Auto,Auto,1,27,19,30,man,N1,8,2015/09/27 20:54,type B,n

A0101,2015/10/01 16:10,1,27,28,Auto,Auto,1,27,19,30,man,N1,8,2015/09/27 20:54,type B,n

A0101,2015/10/01 16:20,1,27,27,Auto,Auto,1,27,19,30,man,N1,8,2015/09/27 20:54,type B,n

A0101,2015/10/01 16:30,1,27,26,Auto,Auto,1,27,19,30,man,N1,8,2015/09/27 20:54,type B,n

A0101,2015/10/01 16:40,1,27,27,Auto,Auto,1,27,19,29,man,N1,8,2015/09/27 20:54,type B,n

A0101,2015/10/01 16:50,1,27,28,Auto,Auto,1,27,19,29,man,N1,8,2015/09/27 20:54,type B,n

A0101,2015/10/01 17:00,1,27,27,Auto,Auto,1,27,19,29,man,N1,8,2015/09/27 20:54,type B,n

A0101,2015/10/01 17:10,1,27,28,Auto,Auto,1,27,19,28,man,N1,8,2015/09/27 20:54,type B,n

A0101,2015/10/01 17:20,1,27,27,Auto,Auto,1,27,19,28,man,N1,8,2015/09/27 20:54,type B,n

A0101,2015/10/01 17:30,1,27,27,Auto,Auto,1,27,19,28,man,N1,8,2015/09/27 20:54,type B,n

A0101,2015/10/01 17:40,1,27,27,Auto,Auto,1,27,19,28,man,N1,8,2015/09/27 20:54,type B,n

A0101,2015/10/01 17:50,1,27,27,Auto,Auto,1,27,19,28,man,N1,8,2015/09/27 20:54,type B,n

A0101,2015/10/01 18:00,1,27,28,Auto,Auto,1,27,19,27,man,N1,8,2015/09/27 20:54,type B,n

A0101,2015/10/01 18:10,1,27,27,Auto,Auto,1,27,19,26,man,N1,8,2015/09/27 20:54,type B,n

The above example is only to illustrate the storage format of the customer's historical check-in information, and the meaning of each field is not important, so it will not be described in detail.

2. Model generation

The method and process for the energy saving control server to acquire the customer's check-in information in an active and passive manner and store it as the customer's historical check-in information have been described above.

After acquiring multiple pieces of historical check-in information of a customer, a comfort level model and a return time probability model corresponding to the customer can be generated by using all the obtained historical check-in information.

In the embodiment of the present invention, in order to generate the comfort level model and the return time probability model corresponding to the customer, the historical check-in information of the customer is divided into four types of instances according to the acquisition time and the occupancy state information of the customer's room.

Under normal circumstances, after a customer checks in at a smart hotel, the change in the occupancy status of the customer's room follows the following rules:

After the customer checks in, the PMS system sets the occupancy status of the customer's room to check in state and reports it to the energy-saving control server, so that the energy-saving control server determines that the occupancy state of the customer's room is check in. And before the customer enters the room, the occupancy status of the room remains in the check in state.

Next, when the customer enters the room, the gateway of the room determines that the occupancy state of the room is occupied and reports it to the energy-saving control server, so that the energy-saving control server determines that the occupancy state of the room where the customer is staying is occupied. And until the customer leaves the room, the occupied state of the room remains occupied.

Next, when the customer leaves the room, the gateway of the room determines that the occupancy state of the room is unoccupied and reports it to the energy-saving control server, so that the energy-saving control server determines that the occupancy state of the room where the customer is staying is unoccupied. And until the customer enters the room again, the occupancy state of the room remains unoccupied.

After that, customers will repeatedly enter and exit the room according to actual needs. Every time a customer enters a room, the gateway of the room determines that the occupancy status of the room is occupied and reports it to the energy-saving control server, so that the energy-saving control server determines that the occupancy status of the room where the customer is staying is occupied. Every time a customer leaves a room, the gateway of the room determines that the occupancy state of the room is unoccupied and reports it to the energy-saving control server, so that the energy-saving control server determines that the occupancy state of the room the customer is staying in is unoccupied. Therefore, before the customer checks out, the occupied state of the room will switch between occupied and unoccupied states according to the specific situation of the customer entering and leaving the room.

Finally, when the customer checks out at the front desk of the hotel, the PMS system sets the occupancy status of the customer's room to check out and reports it to the energy-saving control server, so that the energy-saving control server determines that the occupancy status of the customer's room is check out. And the occupancy status of the room remains checkout until a new customer checks in to the room.

It can be seen that during a customer's check-in process, the occupancy state of the customer's room is only in the check-in state at the beginning, and the check-out state at the end, and switches between the occupied and unoccupied states at other times.

In the above four room occupancy states, the energy-saving control server can actively and passively acquire the customer's check-in information and store it as the customer's historical check-in information.

In the present invention, all historical occupancy information obtained by the energy-saving control server in the same room occupancy state in the same period is taken as an example corresponding to the room occupancy state. E.g:

All customer check-in information obtained by the energy-saving control server within a period of time when the room occupancy state remains checked in is determined as an instance corresponding to the room occupancy state check in;

When the occupancy status of the room changes from check in to occupied, the occupancy information of all customers obtained by the energy-saving control server within a period of time when the occupancy status of the room remains occupied is determined as an instance corresponding to the occupied room status;

When the occupancy status of the room changes from occupied to unoccupied, the occupancy information of all customers obtained by the energy-saving control server within a period of time when the room occupancy status is occupied is determined as an instance corresponding to the unoccupied room occupancy status;

When the occupied state of the room becomes occupied again, the energy-saving control server determines the occupancy information of all customers obtained within a period of time when the occupied state of the room is occupied again as another instance corresponding to the occupied state of the room;

And so on.

It can be seen that in the present invention, when the historical occupancy information of customers is divided into four types of instances according to the acquisition time and the occupancy state information of the customer's room, all historical occupancy information of the customer is actually sorted according to the acquisition time, and adjacent And multiple pieces of historical check-in information with the same room occupancy status are determined as an instance corresponding to the room occupancy status.

In the present invention, the comfort level model corresponding to the customer includes a first comfort level model and a second comfort level model, wherein the first comfort level model is generated using the instance corresponding to the room occupancy status check in, and the second comfort level model uses the room occupancy level. Instances corresponding to occupied are generated.

in,

The generation method of the first comfort level model is:

Determine the comfort level corresponding to each instance whose room occupancy status is check in;

All instances in which the room occupancy state is checked in and the comfort level corresponding to each instance are used as the input of the gradient boosting decision tree (GBDT) algorithm, and the first comfort level model corresponding to the customer is obtained after iterative training.

The generation method of the second comfort level model is as follows:

Determine the comfort level for each instance where the room occupancy state is occupied;

All instances in which the room occupancy state is occupied and the comfort level corresponding to each instance are used as the input of the GBDT algorithm, and the second comfort level model corresponding to the customer is obtained after iterative training.

When the energy-saving device includes a thermostat, the state information of the energy-saving function device includes the state information of the thermostat; in this case, the comfort level corresponding to each instance whose room occupancy state is checked in refers to the thermostat that has been maintained for the longest time in the instance. Set temperature; the comfort level corresponding to each instance where the room occupancy state is occupied also refers to the setting temperature of the thermostat that has been maintained for the longest time in that instance.

In the present invention, the method for determining the maintenance time of each thermostat setting temperature in an instance may specifically be: sorting all historical occupancy information in the instance according to the acquisition time, and calculating a plurality of adjacent thermostat setting temperatures with the same temperature. The latest acquisition time in the historical check-in information (that is, the acquisition time of the last customer’s check-in information after sorting by acquisition time) minus the earliest acquisition time (that is, the first customer’s check-in after sorting by acquisition time) information acquisition time), the calculated difference is taken as the maintenance time of the thermostat set temperature.

In the embodiment of the present invention, the return time probability model corresponding to the customer is generated by using the instance corresponding to the room occupancy state unoccupied, and the specific method is as follows:

For each instance whose room occupancy status is unoccupied, the difference between the latest acquisition time and the earliest acquisition time of all historical occupancy information in the instance is taken as the return time corresponding to the instance;

All instances whose room occupancy status is unoccupied and the return time corresponding to each instance are used as the input of the random forest algorithm, and the return time probability model corresponding to the customer is obtained after training.

In the present invention, for each instance in which the room occupancy status is unoccupied, the difference between the latest acquisition time and the earliest acquisition time of all historical occupancy information in the instance is taken as the return time corresponding to the instance, which is essentially to calculate the room occupancy status. The time to remain unoccupied, that is, the length of time between when a customer leaves the room and returns to the room. Other methods can also be used to calculate, for example, using room entry and exit time calculation, which will not be described in detail here.

3. Using the model for energy-saving control

The generation of the first comfort level model, the second comfort level model, and the return time probability model corresponding to the customer has been described above. The principle of using the first comfort level model, the second comfort level model, and the return time probability model corresponding to the customer to perform energy-saving control on the energy-saving equipment in the customer's room will be described below.

Energy saving control when the room occupancy status is check in:

When the room occupancy status is check in, the energy-saving control of the energy-saving devices in the room is divided into two situations:

The first case: the first comfort level model corresponding to the customer is not generated, and the default configuration is used for energy saving control.

When a customer checks in at a smart hotel for the first time, since the historical check-in information corresponding to the customer has not been stored before, the first comfort level model cannot be generated according to the historical check-in information corresponding to the customer at the time of check-in, so the first comfort level model cannot be used. The comfort level model performs energy-saving control of energy-saving equipment in the customer's room.

In this case, the default configuration is used to control the energy saving of the energy-saving equipment in the customer's room. The default configuration should be based on the principle of maximizing energy saving while ensuring the comfort of the customer.

In the embodiment of the present invention, a default configuration example of the thermostat when the room occupancy state is check in is as follows:

2016-03-02～2016-06-01Heat mode setback temperature: 24 degrees Celsius;

2016-06-02～2016-09-01Cool mode setback temperature: 22 degrees Celsius;

 …

As can be seen from the above configuration example, when the occupancy status of the customer's room is check in, between March and June 2016, the heating setting temperature of the thermostat is set to 24 degrees Celsius by default. Between June and September 2016, the thermostat's cooling setting was set to 22 degrees Celsius by default.

The second situation: the first comfort level model has been generated, and the energy-saving equipment is controlled by using the first comfort level model.

When a customer has checked in at a smart hotel for many times, since many pieces of historical check-in information corresponding to the customer have been stored during the previous check-in, when checking in this time, it can be generated directly according to the stored historical check-in information corresponding to the customer. The first comfort level model corresponding to the customer, so that when the occupancy state of the customer's room is checked in, the energy-saving control of the energy-saving equipment in the customer's room is performed by using the first comfort level model corresponding to the customer.

Taking the energy-saving control of the thermostat as an example, the comfort value corresponding to the instance whose room occupancy status is checked in actually refers to the temperature setting value of the thermostat. The energy-saving control process is as follows:

When the energy-saving control server determines that the occupancy status of the customer's room is check in, the energy-saving control server obtains multiple pieces of current check-in information corresponding to the customer (in order to distinguish it from the historical check-in information, the energy-saving control server determines the occupancy status of the customer's room as check in). check in, and the check-in information corresponding to the customer obtained when the thermostat is energy-saving controlled by using the generated first comfort level model is called the current check-in information), and the multiple pieces of current check-in information are regarded as the room occupancy status as check in An instance of , taking the instance as the input of the first comfort level model corresponding to the customer, the comfort level value corresponding to the instance can be obtained, and the thermostat temperature setting in the room where the customer is staying is adjusted to the comfort level value, thereby Realize energy saving control of thermostat.

Energy saving control when the room occupancy status is occupied:

When the occupancy status of the room is occupied, the energy-saving control of the energy-saving equipment in the customer's room is also divided into two situations:

The first case: The second comfort level model corresponding to the customer has not been generated, and the default configuration is used for energy-saving control.

When a customer checks in at a smart hotel for the first time, within the initial period of time after check-in, because the historical occupancy information of the occupied room has not yet been stored, or the stored historical occupancy information of the occupied room is relatively small, it cannot be based on The historical occupancy information corresponding to the customer generates the second comfort level model corresponding to the customer, so it is also impossible to use the second comfort level model corresponding to the customer to perform energy-saving control on the energy-saving equipment in the customer's room.

In this case, the default configuration is used to control the energy-saving equipment in the customer's room. The content of the default configuration can be set according to actual needs. For example, the default configuration when the room occupancy status is check in can be used, which will not be described in detail here.

The second situation: the second comfort level model corresponding to the customer has been generated, and the energy-saving equipment is controlled by using the second comfort level model corresponding to the customer.

When a customer stays in a smart hotel for the first time and has stayed for a long time, or the customer has stayed in a smart hotel many times before, since many pieces of historical check-in information corresponding to the customer have been stored at this time, the customer can directly The stored historical occupancy information generates the second comfort level model corresponding to the customer, so that when the occupancy status of the customer's room is occupied, the energy-saving equipment in the customer's room is controlled by using the second comfort level model corresponding to the customer.

Taking the energy-saving control of the thermostat as an example, the comfort value corresponding to the instance in which the room occupancy state is occupied actually refers to the temperature setting value of the thermostat, and the energy-saving control process is as follows:

After the energy-saving control server determines that the occupancy status of the customer's room is occupied, the energy-saving control server obtains multiple pieces of current occupancy information corresponding to the customer (in order to distinguish it from the historical occupancy information, the energy-saving control server determines the occupancy status of the customer's room as occupied). , and the occupancy information corresponding to the customer obtained when the thermostat is energy-saving controlled by using the generated first comfort level model is called the current occupancy information), and the multiple pieces of current occupancy information are taken as an instance of the room occupancy status as occupied , the instance is used as the input of the second comfort model corresponding to the customer, the comfort value corresponding to the instance can be obtained, and the thermostat temperature setting in the room where the customer is staying is adjusted to the comfort value, so as to realize the control of the constant temperature energy-saving control of the device.

Energy saving control when the room occupancy state is unoccupied:

When the room occupancy status is unoccupied, the energy-saving control of the energy-saving equipment in the room is also divided into two situations:

The first case: The return time probability model corresponding to the customer has not been generated, and the default configuration is used for energy saving control.

When a customer checks in at a smart hotel for the first time, within the initial period of time after check-in, since there is no historical occupancy information with a room occupancy status of unoccupied, or the stored historical occupancy information with a room occupancy status of unoccupied is relatively small, it cannot be based on The historical check-in information corresponding to the customer generates a return time probability model corresponding to the customer. Therefore, the return time probability model cannot be used to perform energy-saving control on the energy-saving equipment in the customer's room.

In this case, the default configuration is used to control the energy-saving equipment in the customer's room. The content of the default configuration can be set according to actual needs.

In the embodiment of the present invention, a default configuration sample of the thermostat when the room occupancy state is unoccupied is as follows:

2016-03-02～2016-06-01Heat mode setback temperature: 26 degrees Celsius, return time: 1 hour;

2016-06-02～2016-09-01Cool mode setback temperature: 30 degrees Celsius, return time: 1 hour;

 …

As can be seen from the above configuration example, when the occupancy status of the customer's room is unoccupied, between March and June 2016, the heating setting temperature of the thermostat is set to 26 degrees Celsius by default, and the return time is 1 hour (leaving room to return in 1 hour). Between June and September 2016, the thermostat's cooling setting was set to 30 degrees Celsius by default, and the return time was 1 hour.

Using the above default configuration, the thermostat setting temperature of the thermostat can be kept as the temperature value in the default configuration within one hour after the user leaves. Since the temperature setting in the default configuration is less different from the outdoor temperature in the corresponding period, it can make Maintaining the thermostat set temperature at this temperature consumes relatively little energy.

The second situation: the return time probability model corresponding to the customer has been generated, and the energy-saving equipment is controlled by using the return time probability model corresponding to the customer.

When a customer stays in a smart hotel for the first time and has stayed for a long time, or if the customer has stayed in a smart hotel for many times, since many pieces of historical check-in information corresponding to the customer have been stored at this time, they can be directly based on the stored information corresponding to the customer. The return time probability model corresponding to the customer is generated from the historical occupancy information of the customer, so that when the occupancy status of the customer's room is unoccupied, the energy-saving equipment in the customer's room is controlled by using the return time probability model corresponding to the customer.

Taking the energy-saving control of the thermostat as an example, the energy-saving state of the thermostat needs to be preset (the energy-saving state information of the thermostat is a value of the thermostat state information, and the energy consumption can be reduced when the thermostat state is set to the energy-saving state. ), then the energy-saving control process when the room occupancy state is unoccupied is as follows:

When the room occupancy status of the room that the customer is staying in becomes unoccupied, obtain multiple pieces of current occupancy information corresponding to the customer, and use the pieces of current occupancy information as an instance of the unoccupied room occupancy status. The return time probability model determines the return time of the customer, and before the customer returns, the thermostat state in the room where the customer stays is adjusted to a preset energy saving state.

In this embodiment of the present invention, in order to determine the customer's return time, it is also necessary to preset a return time probability threshold (threshold) and a return time interval sequence. The value interval of the return time probability threshold is (0, 1), and the return time interval An example of a sequence is: 0.5 hours, 1 hour, 1.5 hours, 2 hours.

The return time of the customer is determined according to the instance and the return time probability model corresponding to the customer. In fact, the confidence level/probability of each return time interval in the preset return time interval sequence is calculated, and the confidence level is higher than and closest to the return time interval is selected. A return time interval of the time probability threshold is used as the return time of the customer.

An example is used to illustrate the method of determining the customer's return time: assuming that the return time probability threshold is 0.8, and according to the customer's corresponding return time probability model, the probability value of returning within 0.5 hours is 0.5, and the probability value of returning within 1 hour is 0.5. is 0.85, the probability value returned within 1.5 hours is 0.95, and the probability value returned within 2 hours is 0.99. Then it can be determined that the probability of returning within 1 hour is 0.85, which is greater than 0.8 and closest to 0.8. Therefore, the return time of the customer is determined as 1 hour, and the state of the thermostat in the customer's room is maintained before the customer returns to the preset energy-saving state.

Energy saving control when the room occupancy state is unoccupied:

When the customer checks out, the room occupancy status of the customer's room changes to check out. Since there is no need to serve the customer at this time, in order to save energy, the thermostat in the customer's room can be directly turned off to stop working. Thus, energy can be saved to the greatest extent possible.

The implementation principle of the present invention is described in detail above. Based on the above-mentioned principle description, the present invention provides an energy-saving control method for a smart hotel, which is described in detail below with reference to FIG. 2 :

Referring to FIG. 2, FIG. 2 is a flowchart of an energy-saving control method for a smart hotel according to an embodiment of the present invention. The method is applied to an energy-saving control server, as shown in FIG. 2, and mainly includes the following steps:

Step 201: Obtain multiple pieces of historical check-in information of a customer, and generate a comfort level model and a return time probability model corresponding to the customer based on all the obtained historical check-in information;

Wherein, the check-in information includes: the customer's check-in room information; the customer's check-in room information includes: room occupancy status, room entry and exit time, and energy-saving equipment status information;

Step 202: When the room occupancy state of the customer's room changes, use the customer's corresponding comfort model and return time probability model to adjust the state of the energy-saving equipment in the customer's room to achieve energy-saving control.

In the method shown in Figure 2,

Each room of a smart hotel is configured with a corresponding gateway;

A network connection is established between the energy-saving control server and the gateway configured in each room in the smart hotel, and a network connection is also established with the property management server PMS system;

The energy-saving control server obtains the room occupancy status of the room from the PMS system and the gateway configured in each room, and obtains the room entry and exit time of the room and the energy-saving device status information of the room from the gateway configured in the room.

In the method shown in Figure 2,

The room occupancy status includes: check in, occupied, unoccupied, and check out;

When the customer checks in, the PMS system sets the room occupancy status of the customer's room to checkin, and reports it to the energy-saving control server; when the customer checks out, sets the room occupancy status of the customer's room to check out, and reports To the energy-saving control server;

The gateway configured in each room sets the occupancy status of the room to occupied when the customer enters the room, and reports it to the energy-saving control server. When it determines that the customer leaves the room, the occupancy status of the room is set to unoccupied and Report to the energy saving control server.

In the method shown in Figure 2,

Each room of the smart hotel is equipped with a sensor for sensing customer behavior, and the sensor for sensing customer behavior has a network connection with the gateway configured in the room;

The gateway configured in each room of the smart hotel obtains customer behavior information from sensors used to perceive customer behavior, and determines whether the customer enters the room or leaves the room according to the obtained customer behavior information.

In the method shown in Figure 2,

The sensors for sensing customer behavior include door magnetic sensors and motion sensors;

The gateway configured in each room of the smart hotel obtains customer behavior information from sensors used to sense customer behavior, including:

Through the network connection with the door sensor configured in the room, receive the room door state change information reported by the door sensor when sensing the room door state change in the room;

Through the network connection with the motion sensor configured in the room, receive the customer movement information reported by the motion sensor when it senses that the customer movement information is reported;

The customer behavior information is determined according to the state change of the room door received from the door magnetic sensor and the customer movement information received from the motion sensor.

In the method shown in Figure 2,

The room door state change information includes: state change; the state change includes: closed to open, open to closed;

The customer movement information includes: movement direction and movement distance;

The customer behavior information includes: entering the room and leaving the room;

The gateway configured in each room determines the customer behavior information according to the room door state change information received from the door magnetic sensor and the customer movement information received from the motion sensor:

If the state of the room door of the room changes from closed to open, if the moving direction of the customer is to move into the room and the moving distance indicates that the customer has moved into the room, it is determined that the customer behavior information is entering the room, if the moving direction of the customer is If it moves out of the room and the moving distance indicates that the customer has moved out of the room, it is determined that the customer behavior information is to leave the room;

If the state of the room door of the room changes from open to closed, if the customer's moving direction is to move into the room and the moving distance indicates that the customer has moved into the room, it is determined that the customer's behavior information is entering the room, and if the customer's moving direction is Moving out of the room and the moving distance indicates that the customer has moved out of the room, the customer behavior information is determined to be leaving the room.

In the method shown in Figure 2,

The room door state change information further includes: state change time;

The entry and exit time of the room includes: entry time and departure time;

After the gateway configured in each room determines the customer behavior information according to the received room door state change information and the customer movement information, it further includes: if the customer behavior information indicates that the customer enters the room, set the entry time of the room as the room door state change The state change time in the information; if the customer behavior information indicates that the customer leaves the room, the room's departure time is set as the state change time in the room door state change information.

In the method shown in Figure 2,

Each room in the smart hotel is equipped with at least one energy-saving device, and each energy-saving device establishes a network connection with the gateway configured in the room;

The gateway configured in each room receives the changed state information of the energy-saving device reported by the energy-saving device when sensing the change of its own state information through the network connection with each energy-saving device configured in the room.

In the method shown in Figure 2,

The check-in information of the customer also includes: customer information; the customer information includes: name, gender, age, nationality, check-in time, check-in room number, check-in length, and personal preference;

The energy-saving control controller obtains customer information from the PMS system; the PMS system records the customer information and reports it to the energy-saving control server when the customer checks in.

In the method shown in Figure 2,

The customer's check-in information also includes: external environment information, the external environment information includes: weather information, the weather information includes city ID, temperature, humidity, wind speed, and weather description information;

The energy saving control controller obtains the weather information of the smart hotel location from the application module for providing the weather information.

In the method shown in Figure 2,

The comfort level model corresponding to the customer includes a first comfort level model and a second comfort level model;

The comfort level model corresponding to the customer generated based on all the obtained historical occupancy information includes:

Sort all the acquired historical occupancy information according to the acquisition time, and determine multiple pieces of historical occupancy information that are adjacent and have the same room occupancy status as an instance corresponding to the room occupancy status;

Generate the first comfort level model corresponding to the customer based on all instances where the room occupancy state is check in;

Generate a second comfort model for the customer based on all instances where the room occupancy state is occupied.

In the method shown in Figure 2,

The method for generating the first comfort level model corresponding to the customer based on all instances whose room occupancy status is check in is as follows:

For each instance in which the room occupancy state is check in, determine the comfort level corresponding to the instance;

All instances with the room occupancy status as check in and the comfort value corresponding to each instance are used as the input of the gradient boosting decision tree GBDT, and the first comfort model corresponding to the customer is obtained after iterative training;

The method for generating the second comfort model corresponding to the customer based on all instances with the room occupancy status as occupied is:

For each instance where the room occupancy state is occupied, determine the comfort value corresponding to the instance;

All instances in which the room occupancy state is occupied and the comfort value corresponding to each instance are used as the input of GBDT, and the second comfort model corresponding to the customer is obtained after iterative training.

In the method shown in Figure 2,

The energy-saving device includes a thermostat; the energy-saving device state information includes thermostat state information; the thermostat state information includes a thermostat setting temperature, a thermostat fan mode, a cooling setting temperature, and a heating setting temperature;

The method for determining the comfort level value corresponding to the instance is: determining the thermostat setting temperature that lasts the longest in the instance as the comfort level value corresponding to the instance.

In the method shown in Figure 2,

The difference between the latest acquisition time and the earliest acquisition time of all historical occupancy information with the same thermostat setting temperature in the same instance is determined as the maintaining time of the thermostat setting temperature.

In the method shown in Figure 2,

The return time probability model corresponding to the customer generated based on all the obtained historical check-in information includes:

Sort all the acquired historical occupancy information according to the acquisition time, and determine multiple pieces of historical occupancy information that are adjacent and have the same room occupancy status as an instance corresponding to the room occupancy status;

Generates a return time probabilistic model for this customer based on all instances where the room occupancy state is unoccupied.

In the method shown in Figure 2,

The method to generate the return time probability model corresponding to the customer based on all instances whose room occupancy state is unoccupied is:

For each instance whose room occupancy status is unoccupied, the difference between the latest acquisition time and the earliest acquisition time of all historical occupancy information in the instance is taken as the return time corresponding to the instance;

All instances whose room occupancy status is unoccupied and the return time corresponding to each instance are used as the input of the random forest algorithm, and the return time probability model corresponding to the customer is obtained after training.

In the method shown in Figure 2,

When the room occupancy state of the customer's room changes, using the comfort model and return time probability model corresponding to the customer to adjust the state of the energy-saving equipment in the customer's room includes:

When the room occupancy status of the customer's room becomes checked in, obtain multiple pieces of current check-in information corresponding to the customer, use the multiple pieces of current check-in information as an instance, and use the instance as the first comfort level model corresponding to the customer , obtain the comfort value corresponding to the instance, and adjust the thermostat temperature setting of the customer's room to the comfort value;

When the room occupancy status of the customer's room becomes occupied, obtain multiple pieces of current occupancy information corresponding to the customer, use the multiple pieces of current occupancy information as an instance, and use the instance as the second comfort level model corresponding to the customer. Input, get the comfort value corresponding to the instance, and adjust the thermostat temperature setting of the customer's room to the comfort value;

When the room occupancy status of the customer's room becomes unoccupied, obtain multiple pieces of current occupancy information corresponding to the customer, use the multiple pieces of current occupancy information as an instance, and determine the return time probability model corresponding to the instance and the customer. The customer's return time, before the customer returns, adjust the thermostat state of the customer's room to the preset energy-saving state;

When the room occupancy status of the room where the customer is staying is changed to check out, the thermostat in the room where the customer is staying is turned off.

In the method shown in Figure 2,

Preset return time probability threshold and return time interval sequence;

The method for determining the return time of the customer according to the instance and the return time probability model corresponding to the customer is:

Input the instance and each return time interval in the preset return time interval sequence into the return time probability model corresponding to the client to obtain the confidence value corresponding to the return time interval;

The return time interval of which the confidence value is higher than and closest to the preset return time probability threshold is determined as the return time of the customer.

In the method shown in Figure 2,

The method for acquiring multiple pieces of historical occupancy information of the customer is: periodically obtain the current occupancy information of the customer from the PMS system, the gateway of the customer's room, and the local, and receive the PMS system, the gateway of the customer's room, and/or the local module Determine the current check-in information reported when the current check-in information of the client changes, and store the acquired current check-in information as the client's historical check-in information.

In the method shown in Figure 2,

Before storing the acquired current occupancy information as historical occupancy information, it further includes:

If the value of any item in the current check-in information of the customer exceeds the value range of the item, delete the current check-in information of the customer;

If there is a vacancy item in the current check-in information of the customer, fill the vacancy item with the value or the default value of the item in the current check-in information of the customer obtained most recently, and store the current check-in information that fills the vacancy item as the customer's historical occupancy information;

If the above two conditions are not met, the current check-in information of the customer is stored as the historical check-in information of the customer.

The present invention also provides an energy-saving control system for an intelligent hotel, which is described below with reference to the above-mentioned FIG. 1 . As shown in FIG. 1 , the system includes an energy-saving control server;

The energy-saving control server is used for acquiring multiple pieces of historical check-in information of a customer, and generating a comfort level model and a return time probability model corresponding to the customer based on all the obtained historical check-in information; the check-in information includes: the room check-in information of the customer; The room occupancy information of the customer includes: room occupancy status, room entry and exit time, and energy-saving equipment status information; when the room occupancy status of the customer's room changes, the comfort level model corresponding to the customer and the return time probability model are used to compare the data. The state of the energy-saving equipment in the customer's room is adjusted to achieve energy-saving control.

In the system shown in Figure 1,

Each room of a smart hotel is configured with a corresponding gateway;

The system also includes the property management server PMS system and the gateway configured in each room of the smart hotel;

A network connection is established between the energy-saving control server and the gateway configured in each room in the smart hotel, and a network connection is also established with the property management server PMS system;

The energy saving control server obtains the room occupancy status of the room from the PMS system and the gateway configured in each room, and obtains the room entry and exit time of the room and the energy saving device status information of the room from the gateway configured in the room.

In the system shown in Figure 1,

The room occupancy status includes: check in, occupied, unoccupied, and check out;

When the customer checks in, the PMS system sets the room occupancy status of the customer's room to checkin, and reports it to the energy-saving control server; when the customer checks out, sets the room occupancy status of the customer's room to check out, and reports To the energy-saving control server;

The gateway configured in each room sets the occupancy status of the room to occupied when the customer enters the room, and reports it to the energy-saving control server. When it determines that the customer leaves the room, the occupancy status of the room is set to unoccupied and Report to the energy saving control server.

In the system shown in Figure 1,

Each room of a smart hotel is equipped with sensors for sensing customer behavior;

The system also includes sensors configured in each room of the smart hotel to sense customer behavior;

The sensor configured in each room for sensing customer behavior establishes a network connection with the gateway configured in the room;

The gateway configured in each room of the smart hotel obtains customer behavior information from sensors used to perceive customer behavior, and determines whether the customer enters the room or leaves the room according to the obtained customer behavior information.

In the system shown in Figure 1,

The sensors for sensing customer behavior include door magnetic sensors and motion sensors;

The gateway configured in each room of the smart hotel obtains customer behavior information from sensors used to sense customer behavior, including:

Through the network connection with the door sensor configured in the room, receive the room door state change information reported by the door sensor when sensing the room door state change in the room;

Through the network connection with the motion sensor configured in the room, receive the customer movement information reported by the motion sensor when it senses that the customer movement information is reported;

The customer behavior information is determined according to the state change of the room door received from the door magnetic sensor and the customer movement information received from the motion sensor.

In the system shown in Figure 1,

The room door state change information includes: state change; the state change includes: closed to open, open to closed;

The customer movement information includes: movement direction and movement distance;

The customer behavior information includes: entering the room and leaving the room;

When the gateway configured in each room determines the customer behavior information according to the room door state change information received from the door sensor and the customer movement information received from the motion sensor, it is used to:

If the state of the room door of the room changes from closed to open, if the moving direction of the customer is to move into the room and the moving distance indicates that the customer has moved into the room, it is determined that the customer behavior information is entering the room, if the moving direction of the customer is If it moves out of the room and the moving distance indicates that the customer has moved out of the room, it is determined that the customer behavior information is to leave the room;

If the state of the room door of the room changes from open to closed, if the customer's moving direction is to move into the room and the moving distance indicates that the customer has moved into the room, it is determined that the customer's behavior information is entering the room, and if the customer's moving direction is Moving out of the room and the moving distance indicates that the customer has moved out of the room, the customer behavior information is determined to be leaving the room.

In the system shown in Figure 1,

The room door state change information further includes: state change time;

The entry and exit time of the room includes: entry time and departure time;

After the gateway configured in each room determines the customer behavior information according to the received room door state change information and customer movement information, it is further used to: if the customer behavior information indicates that the customer enters the room, set the entry time of the room as the room door state The state change time in the change information; if the customer behavior information indicates that the customer leaves the room, the leaving time of the room is set as the state change time in the room door state change information.

In the system shown in Figure 1,

Each room in a smart hotel is equipped with at least one energy-saving device;

The system also includes energy-saving equipment configured in each room in the smart hotel;

Each energy-saving device configured in each room in the smart hotel establishes a network connection with the gateway configured in the room;

The gateway configured in each room receives the changed state information of the energy-saving device reported by the energy-saving device when sensing the change of its own state information through the network connection with each energy-saving device configured in the room.

In the system shown in Figure 1,

The check-in information of the customer also includes: customer information; the customer information includes: name, gender, age, nationality, check-in time, check-in room number, check-in length, and personal preference;

The PMS system records customer information when the customer checks in and reports it to the energy-saving control server;

The energy-saving control controller receives the customer information entered and reported by the PMS system.

In the system shown in Figure 1,

The customer's check-in information also includes: external environment information, the external environment information includes: weather information, the weather information includes city ID, temperature, humidity, wind speed, and weather description information;

The energy saving control controller obtains the weather information of the smart hotel location from the application module for providing the weather information.

In the system shown in Figure 1,

The comfort level model corresponding to the customer includes a first comfort level model and a second comfort level model;

The comfort level model corresponding to the customer generated by the energy-saving control server based on all the acquired historical occupancy information includes:

Sort all the acquired historical occupancy information according to the acquisition time, and determine multiple pieces of historical occupancy information that are adjacent and have the same room occupancy status as an instance corresponding to the room occupancy status;

Generate the first comfort level model corresponding to the customer based on all instances where the room occupancy state is check in;

Generate a second comfort model for the customer based on all instances where the room occupancy state is occupied.

In the system shown in Figure 1,

When the energy-saving control server generates the first comfort level model corresponding to the client based on all instances whose room occupancy status is checked in, it is used for:

For each instance in which the room occupancy state is check in, determine the comfort level corresponding to the instance;

All instances with the room occupancy status as check in and the comfort value corresponding to each instance are used as the input of the gradient boosting decision tree GBDT, and the first comfort model corresponding to the customer is obtained after iterative training;

When the energy-saving control server generates the second comfort level model corresponding to the customer based on all instances in which the room occupancy state is occupied, it is used for:

For each instance where the room occupancy state is occupied, determine the comfort value corresponding to the instance;

All instances in which the room occupancy state is occupied and the comfort value corresponding to each instance are used as the input of GBDT, and the second comfort model corresponding to the customer is obtained after iterative training.

In the system shown in Figure 1,

The energy-saving device includes a thermostat; the energy-saving device state information includes thermostat state information; the thermostat state information includes a thermostat setting temperature, a thermostat fan mode, a cooling setting temperature, and a heating setting temperature;

When determining the comfort level value corresponding to the instance, the energy saving control server is used for: determining the thermostat setting temperature that lasts the longest in the instance as the comfort level value corresponding to the instance.

In the system shown in Figure 1,

The energy saving control server determines the difference between the latest acquisition time and the earliest acquisition time of all historical occupancy information with the same thermostat set temperature in the same instance as the maintenance time of the thermostat set temperature.

In the system shown in Figure 1,

The energy-saving control server generates a return time probability model corresponding to the customer based on all the acquired historical check-in information, including:

Sort all the acquired historical occupancy information according to the acquisition time, and determine multiple pieces of historical occupancy information that are adjacent and have the same room occupancy status as an instance corresponding to the room occupancy status;

Generates a return time probability model corresponding to this customer based on all instances where the room occupancy state is unoccupied.

In the system shown in Figure 1,

When the energy-saving control server generates the return time probability model corresponding to the customer based on all instances in which the room occupancy state is unoccupied, it is used for:

For each instance whose room occupancy status is unoccupied, the difference between the latest acquisition time and the earliest acquisition time of all historical occupancy information in the instance is taken as the return time corresponding to the instance;

All instances whose room occupancy status is unoccupied and the return time corresponding to each instance are used as the input of the random forest algorithm, and the return time probability model corresponding to the customer is obtained after training.

In the system shown in Figure 1,

The energy-saving control server, when the occupancy state of the room where the customer is staying, changes the state of the energy-saving equipment in the room where the customer is staying by using the comfort level model and the return time probability model corresponding to the customer to adjust the state including:

When the room occupancy status of the customer's room becomes checked in, obtain multiple pieces of current check-in information corresponding to the customer, use the multiple pieces of current check-in information as an instance, and use the instance as the first comfort level model corresponding to the customer , obtain the comfort value corresponding to the instance, and adjust the thermostat temperature setting of the customer's room to the comfort value;

When the room occupancy status of the customer's room becomes occupied, obtain multiple pieces of current occupancy information corresponding to the customer, use the multiple pieces of current occupancy information as an instance, and use the instance as the second comfort level model corresponding to the customer. Input, get the comfort value corresponding to the instance, and adjust the thermostat temperature setting of the customer's room to the comfort value;

When the room occupancy status of the customer's room becomes unoccupied, obtain multiple pieces of current occupancy information corresponding to the customer, use the multiple pieces of current occupancy information as an instance, and determine the return time probability model corresponding to the instance and the customer. The customer's return time, before the customer returns, adjust the thermostat state of the customer's room to the preset energy-saving state;

When the room occupancy status of the room where the customer is staying is changed to check out, the thermostat in the room where the customer is staying is turned off.

In the system shown in Figure 1,

Preset return time probability threshold and return time interval sequence;

When the energy saving control server determines the return time of the client according to the instance and the return time probability model corresponding to the client, it is used for:

Input the instance and each return time interval in the preset return time interval sequence into the return time probability model corresponding to the client to obtain the confidence value corresponding to the return time interval;

The return time interval of which the confidence value is higher than and closest to the preset return time probability threshold is determined as the return time of the customer.

In the system shown in Figure 1,

When the energy-saving control server acquires multiple pieces of historical occupancy information of the customer, it is used for: periodically acquiring the current occupancy information of the customer from the PMS system, the gateway of the customer's room, and the local, and receiving the PMS system and the gateway of the customer's room. , and/or the local module determines the current check-in information reported when the current check-in information of the client changes, and stores the acquired current check-in information as the client's historical check-in information.

In the system shown in Figure 1,

The energy saving control server stores the acquired current occupancy information as historical occupancy information, and is further used for:

If the value of any item in the current check-in information of the customer exceeds the value range of the item, delete the current check-in information of the customer;

If there is a vacancy item in the current check-in information of the customer, fill the vacancy item with the value or the default value of the item in the current check-in information of the customer obtained most recently, and store the current check-in information that fills the vacancy item as the customer's historical occupancy information;

If the above two conditions are not met, the current check-in information of the customer is stored as the historical check-in information of the customer.

As can be seen from the above content, in the present invention, by acquiring massive customer information, customer room information, and external environment information, a corresponding comfort level model and return time probability model are generated for customers, so as to predict comfort level, customer return Time provides precise, personalized service. At the same time, in the present invention, when generating the comfort level model and return time probability model corresponding to the customer, GBDT and random forest algorithm models are used, so that the generated model has high prediction accuracy, is not easy to overfit, and is suitable for high-dimensional mass data. , and it is insensitive to noise data and has good noise tolerance. Moreover, based on massive data and an accurate prediction model, the present invention sets an efficient energy saving strategy that takes into account both comfort and energy saving, which can not only ensure the comfort of customers, but also save energy to the greatest extent, and can improve customer satisfaction for the hotel. and save costs.

The above descriptions are only preferred embodiments of the present invention, and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included in the present invention. within the scope of protection.

Claims (26)

1. An energy-saving control method of an intelligent hotel is applied to an energy-saving control server, and is characterized in that each room of the intelligent hotel is provided with a corresponding gateway which establishes network connection with the energy-saving control server, and the energy-saving control server also establishes network connection with a property management server PMS system; the method comprises the following steps:

acquiring a plurality of pieces of historical check-in information of a client, and generating a comfort level model and a return time probability model corresponding to the client based on all the acquired historical check-in information; the check-in information includes: room check-in information of the client; the client presence information includes: room occupancy state, room entry and exit time, and energy saving device state information; the room occupancy state includes: check in, occupied, unoccupied, check out;

when the room occupancy state of the client entering the room changes, the state of the energy-saving equipment of the client entering the room is adjusted by using the comfort model and the return time probability model corresponding to the client, so as to realize energy-saving control;

wherein,

the energy-saving control server acquires the room occupancy state of the room from the PMS and the gateway configured for each room, and acquires the room entry and exit time of the room and the energy-saving equipment state information of the room from the gateway configured for the room;

when a client registers to live, the PMS sets the room occupation state of the room where the client live to check in and reports the room occupation state to the energy-saving control server; when a client leaves a room, setting the room occupation state of the client entering the room as checkout, and reporting the checkout to an energy-saving control server;

and when determining that the client leaves the room, the gateway configured in each room sets the occupancy state of the room as unicipied and reports the occupancy state to the energy-saving control server.

2. The method of claim 1,

each room of the intelligent hotel is provided with a sensor for sensing the behavior of a customer, and the sensor for sensing the behavior of the customer is connected with a gateway configured in the room through a network;

the gateway configured in each room of the intelligent hotel acquires the customer behavior information from the sensor for sensing the customer behavior, and determines whether the customer enters the room or leaves the room according to the acquired customer behavior information.

3. The method of claim 1,

at least one energy-saving device is configured in each room in the intelligent hotel, and each energy-saving device is in network connection with a gateway configured in the room;

and the gateway configured in each room receives the state information of the changed energy-saving equipment reported by the energy-saving equipment when the energy-saving equipment senses the state information change of the energy-saving equipment through network connection with the energy-saving equipment configured in the room.

4. The method of claim 1,

the check-in information of the client further comprises: customer information; the customer information includes: name, gender, age, nationality, check-in time, check-in room number, check-in duration, personal preferences;

the energy-saving control controller acquires client information from a PMS (permanent magnet system); and the PMS inputs customer information when the customer registers and stays in and reports the customer information to the energy-saving control server.

5. The method of claim 1,

the check-in information of the client further comprises: external environment information, the external environment information including: weather information, wherein the weather information comprises city ID, temperature, humidity, wind speed and weather description information;

and the energy-saving control controller obtains the weather information of the intelligent hotel from the application program module for providing the weather information.

6. The method of claim 3,

the comfort level model corresponding to the client comprises a first comfort level model and a second comfort level model;

generating a comfort model corresponding to the client based on all the acquired historical living-in information includes:

sequencing all the acquired historical check-in information according to acquisition time, and determining a plurality of pieces of historical check-in information which are adjacent and have the same room occupancy state as an example corresponding to the room occupancy state;

generating a first comfort model corresponding to the client for all the instances of check in based on the room occupation state;

and generating a second comfort model corresponding to the client for all the instances of the occupied based on the room occupancy state.

7. The method of claim 6,

the method for generating the first comfort model corresponding to the client for all the instances of check in based on the room occupancy state comprises the following steps:

for each instance of which the room occupancy state is check in, determining a comfort value corresponding to the instance;

taking all the instances with the room occupancy states of check in and the comfort value corresponding to each instance as the input of a gradient lifting decision tree GBDT, and obtaining a first comfort model corresponding to the client after iterative training;

the method for generating the second comfort model corresponding to the client based on all the instances with the room occupied state as the occupied comprises the following steps:

for each instance of which the room occupancy state is occupied, determining a comfort value corresponding to the instance;

and taking all the instances with the room occupation state of occupied and the comfort value corresponding to each instance as the input of the GBDT, and obtaining a second comfort model corresponding to the client after iterative training.

8. The method of claim 7,

the energy saving device comprises a thermostat; the energy saving device state information comprises thermostat state information; the thermostat state information comprises a thermostat setting temperature, a thermostat fan mode, a refrigeration setting temperature and a heating setting temperature;

the method for determining the comfort value corresponding to the example comprises the following steps: the thermostat setting temperature that was maintained for the longest time in this example is determined to be the comfort value for this example.

9. The method of claim 8,

the difference between the latest acquisition time and the earliest acquisition time of all the historical check-in information having the same thermostat setting temperature in the same example is determined as the maintenance time of the thermostat setting temperature.

10. The method of claim 7,

generating a return time probability model corresponding to the client based on all the acquired historical check-in information comprises the following steps:

sequencing all the acquired historical check-in information according to acquisition time, and determining a plurality of pieces of historical check-in information which are adjacent and have the same room occupancy state as an example corresponding to the room occupancy state;

and generating a return time probability model corresponding to the client based on all the instances with the room occupation states of the unicoccuped.

11. The method of claim 10,

the method for generating the return time probability model corresponding to the client based on all the instances with the room occupation states of unicoccuped comprises the following steps:

for each instance with the room occupation state of unicoccured, taking the difference value between the latest acquisition time and the earliest acquisition time of all historical occupancy information in the instance as the return time corresponding to the instance;

and taking all the instances with the room occupation state of unicopied and the return time corresponding to each instance as the input of a random forest algorithm, and obtaining a return time probability model corresponding to the client after training.

12. The method of claim 10,

when the room occupancy state of the client entering the room changes, the adjusting the state of the energy-saving equipment of the client entering the room by using the comfort model and the return time probability model corresponding to the client comprises the following steps:

when the room occupancy state of the room where the client stays is changed into check in, acquiring a plurality of pieces of current stay information corresponding to the client, taking the plurality of pieces of current stay information as an example, taking the example as the input of a first comfort level model corresponding to the client, acquiring a comfort level value corresponding to the example, and adjusting the temperature setting of a thermostat of the client in the room to the comfort level value;

when the room occupancy state of the room where the client stays is changed into occupied, acquiring a plurality of pieces of current stay information corresponding to the client, taking the plurality of pieces of current stay information as an example, taking the example as the input of a second comfort level model corresponding to the client, acquiring a comfort level value corresponding to the example, and adjusting the temperature setting of a thermostat of the client in the room to the comfort level value;

when the room occupancy state of the client living room is changed into unicocculated, acquiring a plurality of pieces of current living information corresponding to the client, taking the plurality of pieces of current living information as an example, determining the return time of the client according to the example and a return time probability model corresponding to the client, and adjusting the state of a thermostat in the client living room to be a preset energy-saving state before the client returns;

when the room occupancy status of the client check-in room changes to check out, the thermostat in the client check-in room is turned off.

13. The method of claim 12,

presetting a return time probability threshold and a return time interval sequence;

the method for determining the return time of the client according to the example and the return time probability model corresponding to the client comprises the following steps:

inputting each return time interval in the example and the preset return time interval sequence into a return time probability model corresponding to the client to obtain a confidence value corresponding to the return time interval;

and determining the return time interval with the confidence value higher than and closest to the preset return time probability threshold as the return time of the client.

14. An intelligent hotel energy-saving control system is characterized by comprising an energy-saving control server, a property management server PMS system which establishes network connection with the energy-saving control server, and a gateway configured for each room of an intelligent hotel;

the energy-saving control server is used for acquiring a plurality of pieces of historical check-in information of the client, and generating a comfort model and a return time probability model corresponding to the client based on all the acquired historical check-in information; the check-in information includes: room check-in information of the client; the client presence information includes: room occupancy state, room entry and exit time, and energy saving device state information; the room occupancy state includes: check in, occupied, unoccupied, check out; when the room occupancy state of the client entering the room changes, the state of the energy-saving equipment of the client entering the room is adjusted by using the comfort model and the return time probability model corresponding to the client, so that energy-saving control is realized;

wherein,

the energy-saving control server acquires the room occupancy state of the room from the PMS and the gateway configured for each room, and acquires the room entry and exit time of the room and the energy-saving equipment state information of the room from the gateway configured for the room;

when a client registers to live, the PMS sets the room occupation state of the room where the client live to check in and reports the room occupation state to the energy-saving control server; when a client leaves a room, setting the room occupation state of the client entering the room as checkout, and reporting the checkout to an energy-saving control server;

and when determining that the client leaves the room, the gateway configured in each room sets the occupancy state of the room as unicipied and reports the occupancy state to the energy-saving control server.

15. The system of claim 14,

each room of the intelligent hotel is provided with a sensor for sensing the behavior of a customer;

the system also comprises a sensor configured for each room of the intelligent hotel for sensing customer behavior;

the sensor for sensing the customer behavior of each room configuration is connected with the gateway of the room configuration through a network;

the gateway configured in each room of the intelligent hotel acquires the customer behavior information from the sensor for sensing the customer behavior, and determines whether the customer enters the room or leaves the room according to the acquired customer behavior information.

16. The system of claim 14,

at least one energy-saving device is configured in each room in the intelligent hotel;

the system also comprises energy-saving equipment arranged in each room in the intelligent hotel;

each energy-saving device configured in each room in the intelligent hotel establishes network connection with a gateway configured in the room;

and the gateway configured in each room receives the state information of the changed energy-saving equipment reported by the energy-saving equipment when the energy-saving equipment senses the state information change of the energy-saving equipment through network connection with the energy-saving equipment configured in the room.

17. The system of claim 14,

the check-in information of the client further comprises: customer information; the customer information includes: name, gender, age, nationality, check-in time, check-in room number, check-in duration, personal preferences;

the PMS inputs customer information when a customer registers and stays in and reports the customer information to the energy-saving control server;

and the energy-saving control controller receives the customer information which is input and reported by the PMS.

18. The system of claim 14,

the check-in information of the client further comprises: external environment information, the external environment information including: weather information, wherein the weather information comprises city ID, temperature, humidity, wind speed and weather description information;

and the energy-saving control controller obtains the weather information of the intelligent hotel from the application program module for providing the weather information.

19. The system of claim 16,

the comfort level model corresponding to the client comprises a first comfort level model and a second comfort level model;

the energy-saving control server generates a comfort model corresponding to the client based on all the acquired historical check-in information, and the comfort model comprises the following steps:

sequencing all the acquired historical check-in information according to acquisition time, and determining a plurality of pieces of historical check-in information which are adjacent and have the same room occupancy state as an example corresponding to the room occupancy state;

generating a first comfort model corresponding to the client for all the instances of check in based on the room occupation state;

and generating a second comfort model corresponding to the client for all the instances of the occupied based on the room occupancy state.

20. The system of claim 19,

when the energy-saving control server generates the first comfort level model corresponding to the client for all the instances of check in based on the room occupancy state, the energy-saving control server is used for:

for each instance of which the room occupancy state is check in, determining a comfort value corresponding to the instance;

taking all the instances with the room occupancy states of check in and the comfort value corresponding to each instance as the input of a gradient lifting decision tree GBDT, and obtaining a first comfort model corresponding to the client after iterative training;

when the energy-saving control server generates a second comfort model corresponding to the client for all the instances of the occupied based on the room occupancy state, the energy-saving control server is configured to:

for each instance of which the room occupancy state is occupied, determining a comfort value corresponding to the instance;

and taking all the instances with the room occupation state of occupied and the comfort value corresponding to each instance as the input of the GBDT, and obtaining a second comfort model corresponding to the client after iterative training.

21. The system of claim 20,

the energy saving device comprises a thermostat; the energy saving device state information comprises thermostat state information; the thermostat state information comprises a thermostat setting temperature, a thermostat fan mode, a refrigeration setting temperature and a heating setting temperature;

when the energy-saving control server determines the comfort value corresponding to the instance, the energy-saving control server is configured to: the thermostat setting temperature that was maintained for the longest time in this example is determined to be the comfort value for this example.

22. The system of claim 21,

and the energy-saving control server determines the difference value between the latest acquisition time and the earliest acquisition time of all historical check-in information with the same thermostat setting temperature in the same example as the maintenance time of the thermostat setting temperature.

23. The system of claim 20,

the energy-saving control server generates a return time probability model corresponding to the client based on all the acquired historical check-in information, and the return time probability model comprises the following steps:

sequencing all the acquired historical check-in information according to acquisition time, and determining a plurality of pieces of historical check-in information which are adjacent and have the same room occupancy state as an example corresponding to the room occupancy state;

and generating a return time probability model corresponding to the client based on all the instances with the room occupation states of the unicoccuped.

24. The system of claim 23,

when the energy-saving control server generates the return time probability model corresponding to the client based on all the instances of which the room occupation states are unicuspid, the energy-saving control server is used for:

for each instance with the room occupation state of unicoccured, taking the difference value between the latest acquisition time and the earliest acquisition time of all historical occupancy information in the instance as the return time corresponding to the instance;

and taking all the instances with the room occupation state of unicopied and the return time corresponding to each instance as the input of a random forest algorithm, and obtaining a return time probability model corresponding to the client after training.

25. The system of claim 23,

when the room occupancy state of the client living room changes, the energy-saving control server adjusts the state of the energy-saving equipment of the client living room by using the comfort model and the return time probability model corresponding to the client, and comprises the following steps:

when the room occupancy state of the room where the client stays is changed into check in, acquiring a plurality of pieces of current stay information corresponding to the client, taking the plurality of pieces of current stay information as an example, taking the example as the input of a first comfort level model corresponding to the client, acquiring a comfort level value corresponding to the example, and adjusting the temperature setting of a thermostat of the client in the room to the comfort level value;

when the room occupancy state of the room where the client stays is changed into occupied, acquiring a plurality of pieces of current stay information corresponding to the client, taking the plurality of pieces of current stay information as an example, taking the example as the input of a second comfort level model corresponding to the client, acquiring a comfort level value corresponding to the example, and adjusting the temperature setting of a thermostat of the client in the room to the comfort level value;

when the room occupancy state of the client living room is changed into unicocculated, acquiring a plurality of pieces of current living information corresponding to the client, taking the plurality of pieces of current living information as an example, determining the return time of the client according to the example and a return time probability model corresponding to the client, and adjusting the state of a thermostat in the client living room to be a preset energy-saving state before the client returns;

when the room occupancy status of the client check-in room changes to check out, the thermostat in the client check-in room is turned off.

26. The system of claim 25,

presetting a return time probability threshold and a return time interval sequence;

when the energy-saving control server determines the return time of the client according to the instance and the return time probability model corresponding to the client, the energy-saving control server is used for:

inputting each return time interval in the example and the preset return time interval sequence into a return time probability model corresponding to the client to obtain a confidence value corresponding to the return time interval;

and determining the return time interval with the confidence value higher than and closest to the preset return time probability threshold as the return time of the client.

Images