A hotel energy saving and guest room operation and maintenance method and system based on people perception

Abstract

The present application belongs to the technical field of hotel intelligent energy-saving, and specifically relates to a hotel energy-saving and guest room operation and maintenance method and system based on number perception, wherein the method comprises the following steps: detecting the direction of personnel entering and exiting and accumulating the number of personnel through a non-vision sensor array installed in a porch; an edge computing gateway calculates the real-time number of personnel in a room by using a number state machine algorithm in combination with PMS registration of the number of personnel, and then generates a multi-strategy control instruction; and a smart guest control host executes differentiated energy-saving control on devices such as air conditioners, lighting, sockets and fresh air, and simultaneously realizes operation and maintenance management such as over-occupancy early warning, cleaning scheduling optimization and illegal occupancy detection. The present application realizes accurate number perception and multi-system linkage under the premise of guaranteeing the privacy of guests, and significantly improves the hotel energy-saving efficiency and operation and management level.

Description

Technical Field

[0001] This invention relates to the field of intelligent energy-saving technology for hotels, and in particular to a method and system for hotel energy saving and guest room operation and maintenance based on human presence perception. Background Technology

[0002] The card-operated power system requires guests to insert their room cards to connect to the main power supply in the room, and then disconnects the power after a delay when the card is removed. This system cannot distinguish the actual number of people in the room. When only one person leaves a room occupied by multiple people, removing the card will cause a power outage in the area where the person left is staying, severely impacting the user experience. If the card is not removed, all electrical appliances in the room will continue to consume power, resulting in energy waste.

[0003] Infrared human body sensing technology detects human movement by installing passive infrared sensors on the ceiling or walls, automatically cutting off power when no one is moving. However, passive infrared sensors have an inherent flaw of false alarms when guests are stationary. When guests are asleep or quietly reading, they are easily misjudged as unoccupied, leading to the accidental shutdown of equipment such as air conditioning and lighting, severely reducing the comfort of the stay.

[0004] Camera image recognition solutions use indoor cameras to collect images and count the number of people. Hotel rooms are highly private spaces, and installing cameras there poses serious privacy compliance risks. Guests are strongly opposed to such solutions, making large-scale application in commercial environments difficult.

[0005] Existing doorway counting technology uses infrared beams or 3D vision devices to count the number of people at the door in some classroom and office building scenarios. However, such solutions only solve the basic counting problem and fail to form a deep linkage control with the hotel's unique guest control system, such as PMS system, door locks, power switches, air conditioning, lighting and fresh air system. Furthermore, they do not design multi-dimensional management strategies for hotel room scenarios, such as overcrowding warning, intelligent judgment of cleaning needs, and detection of unauthorized occupancy. Summary of the Invention

[0006] This invention proposes a hotel energy-saving and room operation and maintenance method and system based on the number of guests, aiming to solve the technical problem in the prior art that it is impossible to accurately sense the number of guests in hotel rooms in real time while protecting the privacy of guests, and to control the electrical equipment and room operation and maintenance management based on the sensing results.

[0007] In a first aspect, the present invention provides a hotel energy-saving and guest room operation and maintenance method based on human presence perception, comprising:

[0008] The direction of people passing through the porch is detected by an array of non-visual sensors installed in the guest room porch area, and the number of people entering and leaving is counted separately.

[0009] The edge computing gateway receives the number of people entering and leaving, and, combined with the initial number of occupants registered in the PMS system, calculates the real-time number of people in the room using a people state machine algorithm;

[0010] Based on the real-time number of people in the room, generate and issue multi-strategy control commands;

[0011] The intelligent guest room control host receives and executes the multi-strategy control instructions to perform energy-saving control of electrical equipment in the guest room or to perform guest room operation and maintenance management;

[0012] The number of people state machine algorithm includes:

[0013] The initial room occupancy is set to the initial occupancy number registered in the PMS system;

[0014] When a person enters the room, the number of people in the room increases by 1; when a person leaves the room, the number of people in the room decreases by 1.

[0015] The confidence level of the calculated real-time number of people in the room is verified. When the real-time number of people conflicts with the status of the power switch and the data of the PIR auxiliary sensor, the cross-validation mechanism is activated and the correction is made according to the majority consensus principle.

[0016] The corrected real-time number of people and status changes in the room are pushed to downstream devices.

[0017] Furthermore, the multi-strategy control instructions include dynamic power access permission management instructions, wherein the dynamic power access permission management includes:

[0018] The power switch remains active when the number of people in the room is greater than 0.

[0019] When the number of people in the room is 0 and continues to exceed the first preset threshold, cut off unnecessary power and keep the air conditioner on standby or switch to energy-saving mode.

[0020] Power will be automatically restored when the number of people in the room changes from 0 to greater than 0.

[0021] Furthermore, the multi-strategy control instructions include occupancy adaptive air conditioning control instructions, wherein the occupancy adaptive air conditioning control includes:

[0022] The target temperature and airflow are dynamically adjusted based on the real-time number of people in the room.

[0023] When the number of users in real time is 1, the system will operate according to the standard settings.

[0024] When the number of people in real time is greater than or equal to 2, the target temperature is adjusted according to the average heating and cooling load per person.

[0025] When the number of people in real time is 0, switch to energy-saving and heat preservation mode.

[0026] Furthermore, the multi-strategy control instructions include lighting zone control instructions, wherein the lighting zone control includes:

[0027] Based on the real-time number of people in the room and the direction of entry detected by the doorway sensing module, the activity area of ​​people is predicted, and the lighting equipment in the corresponding area is turned on or off.

[0028] Furthermore, the multi-strategy control command includes a fresh air volume dynamic adjustment command, wherein the fresh air volume dynamic adjustment includes:

[0029] The opening of the fresh air valve is adjusted in real time according to the number of people in the room.

[0030] When the number of people in real time is 0, close or control the fresh air valve to operate at the minimum air volume;

[0031] When the number of people increases in real time, the fresh air volume is increased proportionally.

[0032] Furthermore, the guest room operation and maintenance management includes overcrowding warning management, which includes:

[0033] When the number of people in a room exceeds the preset threshold for the number of people registered in the PMS system, and the duration exceeds the second preset threshold, an overcrowding warning message is pushed to the hotel management terminal.

[0034] Furthermore, the guest room operation and maintenance management includes cleaning demand judgment management, which includes:

[0035] When the number of people in the room changes from greater than 0 to 0, and the door lock is closed and the power switch is off, it is determined that the guest has left the room, and the room cleaning system is pushed with the cleanable status information; the room usage intensity index is estimated based on the historical data of the number of people in the room, and the cleaning priority is dynamically adjusted based on the usage intensity index.

[0036] Furthermore, the guest room operation and maintenance management includes unauthorized check-in detection management, which includes:

[0037] When the door lock is closed and the door is not opened by swiping a card, and the number of people in the room changes from 0 to greater than 0, an illegal intrusion warning is triggered, and the door lock alarm and security system are activated.

[0038] Secondly, the present invention provides a hotel energy-saving and guest room operation and maintenance system based on human presence perception, the system being used to execute the method, including:

[0039] The porch sensing module, installed on the top or side frame of the guest room door, uses a non-visual sensor array to detect the direction, speed and body shape characteristics of people passing through the porch, and generates the number of people entering and leaving.

[0040] An edge computing gateway is communicatively connected to the porch sensing module and has a built-in people state machine algorithm to calculate the real-time number of people in the room based on the number of people entering, leaving, and the initial number of occupants registered in the PMS system, and to execute local policy decisions.

[0041] The intelligent guest room control host is electrically connected to the electrical equipment in the guest room and is used to receive and execute multi-strategy control commands issued by the edge computing gateway.

[0042] The cloud management platform communicates with the edge computing gateway via the Internet of Things protocol to receive real-time data, energy consumption data, and abnormal event data from each guest room, and performs big data analysis and strategy optimization.

[0043] Mobile terminals are provided for use by hotel managers and guests to view room status, receive abnormal alerts, and remotely control guest room equipment.

[0044] The technical effects of this invention are:

[0045] First, the present invention uses a non-visual sensor array installed in the guest room porch area as the sensing front end. Through non-image acquisition devices such as infrared beam sensors, time-of-flight sensors or millimeter-wave radar, it detects the direction and movement characteristics of people passing through the porch. Its technical effect is that it fundamentally avoids the collection of any image or video information in the guest room, completely eliminates guests' concerns about privacy leaks, and at the same time achieves accurate determination of the direction of people entering and leaving.

[0046] Secondly, this invention constructs a state machine algorithm for the number of people, which includes initial value setting, dynamic updating and cross-validation mechanisms. The number of guests registered by the PMS system is used as the initial value. The number of guests is increased or decreased in real time according to the entry or exit events reported by the doorway sensing module. Cross-validation is initiated when the calculation result conflicts with the power switch status and the data of the passive infrared auxiliary sensor. Its technical effect is to significantly improve the robustness and accuracy of the number of people count in complex hotel scenarios, and effectively solve the counting error problem in situations such as multiple people running in parallel, luggage obstructing the view or children following.

[0047] Third, based on the calculated real-time number of people in the room, this invention generates and executes multi-strategy control commands covering dynamic power access management, adaptive air conditioning control based on the number of people, lighting zone control, and dynamic adjustment of fresh air volume. Its technical effect is to maintain room power supply when multiple people are staying and one person is away, avoiding the pain points of traditional card-operated power supply. It achieves deep energy saving when the room is truly empty, with a comprehensive energy saving rate of 25% to 40%. At the same time, it finely adjusts the air conditioning temperature, lighting zone, and fresh air volume according to changes in the number of people to improve energy efficiency and living comfort.

[0048] Fourth, based on the occupancy perception, this invention further generates value-added strategy instructions for hotel management, including overcrowding warning, cleaning demand judgment, and illegal occupancy detection. Its technical effect is to upgrade the occupancy perception data from a single energy-saving control signal to a data source for hotel operation and management, providing hotels with management functions such as preventing illegal group rentals, optimizing cleaning scheduling, and improving security and water safety.

[0049] Fifth, this invention adopts a two-layer architecture of edge computing gateway and cloud management platform. The edge computing gateway is deployed locally to be responsible for real-time data processing and strategy decision-making to ensure the immediacy and reliability of control. The cloud management platform aggregates data from various stores for big data analysis and strategy optimization. Its technical effect is that it supports rapid deployment and low-latency response for individual hotels, and meets the management needs of group chain hotels for multi-store data aggregation, energy consumption pattern analysis and unified operation decision-making. Attached Figure Description

[0050] Figure 1 This is a flowchart illustrating a hotel energy-saving and guest room operation and maintenance method based on human presence perception, as proposed in an embodiment of the present invention. Detailed Implementation

[0051] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0052] To address the technical problem mentioned in the background art of accurately sensing the real-time number of guests in hotel rooms while ensuring guest privacy, and then using this sensing result to control electrical equipment and manage room operations, this embodiment provides a hotel energy-saving and room operation and maintenance method and system based on guest room sensing. Its core lies in detecting the direction of people entering and exiting through a non-visual sensor array installed in the guest room porch area, combining this with the initial occupancy number registered by the PMS system, calculating the real-time number of guests in the room using a guest room state machine algorithm, and generating multi-strategy control commands based on this real-time number of guests to achieve energy-saving control of electrical equipment in the guest rooms and management of the room operation and maintenance status.

[0053] The overall architecture of the system of this invention is as follows: Figure 1 As shown, it includes a porch sensing module, an edge computing gateway, an intelligent guest room control host, a cloud management platform, and a mobile terminal.

[0054] The porch sensing module is installed at the center of the top beam or side frame of the guest room door frame. The module employs a non-visual sensor array, including but not limited to an infrared beam sensor array, a time-of-flight sensor array, a millimeter-wave radar sensor, or a combination thereof. In this embodiment, a combination of a time-of-flight sensor array and an infrared beam sensor array is used. The time-of-flight sensor array emits modulated infrared light downwards, and depth map data of the porch area is obtained by measuring the round-trip time between the emission and reception of the light pulse. The porch sensing module has a built-in MCU microcontroller, which pre-installs an edge AI algorithm. After receiving the depth map data, this edge AI algorithm processes the data locally, identifying the silhouette shape, direction of movement, and body shape characteristics of a person passing through to determine whether the passage is an entry or exit event. Specifically, this edge AI algorithm is a lightweight convolutional neural network or a decision tree-based classifier, pre-trained in the cloud using a large amount of depth map data from porch scenarios. After training, the model parameters are deployed to the MCU of the porch sensing module. When performing direction determination, the MCU inputs a frame of depth map data collected by the time-of-flight sensor into the model. The model outputs the corresponding direction label for that frame of data, including entry, exit, or invalid noise. For invalid noise, such as scenes where suitcase wheels, pets, or large luggage packages pass by without any people passing by, the model marks them as invalid events and does not include them in subsequent counting.

[0055] The infrared beam sensor array serves as a redundant verification channel. When the time-of-flight sensor fails to transmit data due to strong light interference or physical obstruction, the system automatically switches to infrared beam mode. By detecting the sequence of interruptions in the infrared beams, the system determines the direction of movement, thus ensuring the continuity of the count. After completing direction determination and counting, the porch sensing module only uploads the count results—the cumulative number of people entering and leaving the porch—to the edge computing gateway via Zigbee 3.0 wireless protocol or wired RS485 communication. The porch sensing module does not upload any raw depth maps, images, or video stream data, thus providing dual protection for guest privacy at both the physical and data link levels.

[0056] The edge computing gateway is deployed in the floor's low-voltage electrical room or inside guest rooms, maintaining a communication connection with the porch sensing module. The edge computing gateway incorporates a built-in people state machine algorithm, which operates according to the following steps.

[0057] Initialization steps. After the guest room is cleaned and the PMS system processes check-in, the edge computing gateway receives the number of registered guests for that guest room from the PMS system, denoted as Ninitial, and assigns the current real-time number of guests in the room, Nroom, to Ninitial. At this time, the cumulative number of guests entering (Nin) and the cumulative number of guests leaving (Nout) are both reset to zero.

[0058] Event-triggered update steps. When the porch sensing module detects a person passing through and reports an entry event, the edge computing gateway increments the cumulative number of people entering (Nin) by 1, and simultaneously updates the real-time number of people in the room (Nroom) to the current value plus 1. When the porch sensing module reports a departure event, the edge computing gateway increments the cumulative number of people leaving (Nout) by 1, and simultaneously updates the real-time number of people in the room (Nroom) to the current value minus 1. Its core calculation logic is as follows:

[0059] Nroom = Nin − Nout + Ninitial;

[0060] Where Nroom represents the current real-time number of people in the room, Nin represents the total number of people who have entered since initialization, Nout represents the total number of people who have left since initialization, and Ninitial represents the initial number of people registered in the PMS system. All four variables are dimensionless and represent the number of people.

[0061] Confidence verification and cross-validation steps. After updating Nroom, the edge computing gateway initiates a confidence verification process to detect potential missed or over-detections in the doorway sensor. The edge computing gateway receives human movement signals from passive infrared auxiliary sensors in guest rooms and the on / off status of power switches in real time. Cross-validation is triggered when the following conflicting conditions occur.

[0062] Scenario 1: Nroom is calculated to be 0, but the passive infrared sensor in the guest room continues to detect human movement. In this case, the edge computing gateway determines that the doorway counting has missed a detection, meaning someone entered but was not recorded by the doorway sensing module. The system forcibly corrects Nroom to 1 and simultaneously corrects the accumulated values ​​of Nin or Nout to ensure consistency in subsequent counts. At the same time, it pushes a system self-check event to the cloud management platform, indicating that the doorway sensing module for this guest room may need calibration.

[0063] Scenario 2: The calculated Nroom value is greater than 0, but the power switch is off, and this off-state duration exceeds a third preset threshold (default value is 30 minutes). Simultaneously, the passive infrared sensor does not detect any movement signals during this time period. In this case, the edge computing gateway determines that Nroom may have been overcounted, meaning that the person has actually left but was not recorded by the porch sensing module, or the guest has been in a static, sleeping state for an extended period. Instead of automatically correcting Nroom, the system pushes an anomaly marker event to the cloud management platform, allowing hotel management personnel to manually verify the historical entry and exit timeline via mobile devices.

[0064] The cross-validation mechanism adopts the majority consensus principle. When two of the three data points—porch sensing data, power switch status, and passive infrared sensor data—indicate the same state, the edge computing gateway corrects Nroom based on that state and pushes the correction result and state change event to the intelligent guest room control host and the cloud management platform.

[0065] Time window filtering steps. When the porch sensing module detects continuous rapid entry and exit events, such as a waiter briefly delivering items or a guest briefly leaving to retrieve something, the edge computing gateway sets a time window W, with a default value of three minutes. Within the time window W, although Nroom may fluctuate from 1 to 0 and then back to 1, the edge computing gateway does not send every fluctuation within the window to the intelligent guest room control host. Instead, it waits until the window ends and then generates a final control command based on the net change within the window. This filtering mechanism effectively avoids excessively frequent start-stop cycles for equipment such as air conditioning compressors and lighting relays due to frequent changes in the number of people, extending the lifespan of the equipment.

[0066] The intelligent guest room control unit is electrically connected to the air conditioning equipment, lighting equipment, power sockets, fresh air valve actuators, electronic door locks, and power switches in the guest rooms via relays or a communication bus. The intelligent guest room control unit receives multi-policy control commands from the edge computing gateway and executes corresponding energy-saving control and operation and maintenance management operations.

[0067] For energy-saving control, the edge computing gateway generates the following specific control instructions based on the value of Nroom.

[0068] Dynamic power access control commands. When Nroom is greater than 0, the edge computing gateway sends a valid power access command to the smart guest room control host. The smart guest room control host keeps the power switch closed and on, ensuring the main power supply to the guest room is connected. When Nroom equals 0 and this state continues for more than a first preset threshold T1 (the default value of T1 is five minutes), the edge computing gateway sends an energy-saving standby command to the smart guest room control host. Upon receiving this command, the smart guest room control host cuts off non-essential power supplies in the guest room, including but not limited to power outlets and decorative lighting fixtures, leaving only the air conditioning unit in standby mode or switching the air conditioning to energy-saving insulation mode. When a guest returns, the doorway sensing module detects the entry event, causing Nroom to change from 0 to greater than 0. The edge computing gateway immediately sends a power restoration command, and the smart guest room control host automatically closes the power switch and restores all disconnected power circuits. The entire process does not require the guest to insert a card.

[0069] Adaptive air conditioning control commands based on occupancy. The edge computing gateway has built-in logic for calculating the target temperature and airflow of the air conditioner. When Nroom equals 1, the intelligent guest room control unit controls the air conditioner to operate according to the standard temperature value preset by the guest, such as 24 degrees Celsius, and the standard airflow level. When Nroom is greater than or equal to 2, the edge computing gateway fine-tunes the target temperature based on the room area and the average cooling and heating load coefficient per person. Taking a standard double room as an example, when two people are detected in the room, due to the increased heat dissipation from the human body, the edge computing gateway adjusts the cooling target temperature down by 0.5 degrees Celsius to compensate for the heat load of the human body, and at the same time increases the fan speed from medium speed to medium-high speed to accelerate air convection and heat exchange. When Nroom equals 0, the edge computing gateway sends an energy-saving insulation mode command to the intelligent guest room control unit, which controls the air conditioner to switch to the insulation setpoint of 28 degrees Celsius in summer and 18 degrees Celsius in winter, and controls the fan to run intermittently at low speed to maintain the basic room temperature level to avoid excessive energy loss.

[0070] Lighting zone control commands. The edge computing gateway predicts the guest's activity area based on the entry direction information reported by the porch sensing module and the current Nroom value. When a guest enters the room from outside, the edge computing gateway sends a command to the smart guest room control unit to turn on the lighting in the entryway and bathroom areas. After the guest continues to move towards the bedroom area, after a preset delay, such as fifteen seconds, the edge computing gateway sends a command to turn on the bedroom lighting while simultaneously turning off the entryway lighting. When Nroom changes from greater than 0 to 0, the edge computing gateway sends a command to turn off all room lighting, thus achieving refined lighting control where lights turn on when someone enters and off when they leave.

[0071] Dynamic adjustment commands for fresh air volume. The edge computing gateway has a built-in table showing the correspondence between the fresh air valve opening and Nroom. When Nroom equals 0, the edge computing gateway sends a fresh air shut-off command to the intelligent guest room control unit. The intelligent guest room control unit then controls the fresh air valve actuator to rotate to the fully closed position, or maintains a minimum opening of 5% to meet the minimum requirements for building fire safety and hygiene. When Nroom increases, the edge computing gateway increases the fresh air valve opening proportionally. When Nroom is 1, the opening is 40%; when Nroom is 2, the opening is 70%; and when Nroom is greater than or equal to 3, the opening is 100%. This dynamic adjustment ensures that the indoor carbon dioxide concentration matches the number of people, minimizing the energy consumption of fresh air handling while meeting comfort requirements.

[0072] For operation and maintenance management, the edge computing gateway generates the following management policy instructions based on the Nroom value, door lock status, and power switch status.

[0073] Overcrowding Warning Management Instructions. The edge computing gateway monitors the difference between Nroom and Ninitial registered in the PMS system in real time. When Nroom exceeds the sum of Ninitial and a preset overcrowding threshold (default value: 2 people), and the overcrowding status lasts for more than a second preset threshold T2 (default value: 15 minutes), the edge computing gateway pushes a yellow warning message to the hotel front desk PMS terminal, indicating that the room may be overcrowded. When Nroom exceeds the sum of Ninitial and another higher threshold, such as exceeding 4 people, or the overcrowding status lasts for more than 30 minutes, the edge computing gateway upgrades the warning to a red warning and simultaneously pushes it to the security manager's mobile terminal. Hotel management personnel can use the mobile terminal to access the guest room's entry and exit timeline data for the past two hours to help determine whether the overcrowding situation is due to normal visitors staying briefly, illegal group rentals, or unauthorized entry by strangers.

[0074] Cleaning demand assessment and management commands. The edge computing gateway continuously monitors the changing trend of Nroom, the locked / unlocked status of the door lock, and the on / off status of the power switch. When it detects that Nroom changes from greater than 0 to 0, the door lock is closed, and the power switch is off, the edge computing gateway determines that the guest has left the room and the room is ready for cleaning. At this time, it pushes a status information with a "ready to clean" tag to the guest room cleaning scheduling system. Simultaneously, the edge computing gateway calculates the room usage intensity index Iuse based on the stored historical Nroom data for that guest room. The calculation method is as follows:

[0075] Iuse = Σ(Nroom × t) / B;

[0076] Where Nroom represents the real-time number of guests in each time period, dimensionless; t represents the length of the corresponding time period, in hours; B is the baseline value, in the form of guests multiplied by hours, set by the hotel based on the product of the standard occupancy and standard stay duration for that room type, typically 8 hours. For example, the baseline value for a standard double room is 2 people multiplied by 8 hours, which equals 16 person-hours. Therefore, Iuse is a dimensionless ratio, representing the multiple of the actual usage intensity on a given day relative to the standard intensity. The higher the Iuse value, the greater the usage intensity of the room on that day, resulting in more waste and damage. The edge computing gateway pushes Iuse to the cleaning scheduling system, which prioritizes multiple rooms to be cleaned based on Iuse, assigning cleaning staff to rooms with higher Iuse values ​​first. After cleaning is completed, the cleaning staff confirms completion via an application on their mobile terminal. This confirmation instruction is sent to the edge computing gateway via the cloud management platform. The edge computing gateway resets Nroom to Ninitial and clears Nin and Nout to zero, preparing for the next round of occupancy detection.

[0077] Unauthorized check-in detection and management instructions. The edge computing gateway correlates door lock events and occupancy data in real time. When a door lock is detected to be closed, and the unlocking was not triggered by a legitimate card swipe, and Nroom changes from 0 to greater than 0, the edge computing gateway determines the event as unauthorized intrusion and immediately triggers an unauthorized intrusion warning instruction. On one hand, it triggers an audible and visual alarm on the door lock alarm in the guest room via the smart guest room control host; on the other hand, it pushes the warning information to the security system terminal to notify security personnel to verify the situation.

[0078] The cloud management platform maintains communication with edge computing gateways deployed on various floors or in individual guest rooms via IoT protocols such as MQTT. The platform receives real-time Nroom data, energy consumption data, abnormal event data, and equipment operating status data reported by each guest room, and stores, cleans, and aggregates this data for analysis. At the big data analytics level, the platform statistically analyzes occupancy rates for different room types and floors within the same hotel, breaks down energy consumption for air conditioning, lighting, and ventilation by the number of guests, and identifies high-energy-consuming periods and low-efficiency equipment. At the strategy optimization level, the platform iteratively optimizes air conditioning settings for different locations based on historical data and external weather data, and distributes the optimized parameter templates to each edge computing gateway. At the group-wide operation level, the platform supports multi-store data aggregation, generating unified group-wide operational reports for operational decision-makers to analyze occupancy patterns, abnormal event distribution characteristics, and energy-saving performance comparisons across stores, providing data support for equipment selection and upgrades.

[0079] Mobile terminals include management terminals used by hotel managers and service terminals used by guests. Managers can use management terminals to view the current Nroom status of each guest room in real time, receive overcrowding alerts, unauthorized intrusion alerts, and push notifications about abnormal devices. They can also remotely view the historical entry and exit timeline of specific rooms to assist in incident tracing and arbitration. Guests can use service terminals, such as the hotel's WeChat mini-program or dedicated application, to view the current operating status of devices in their room and, with legal authorization, remotely adjust the target temperature of the air conditioning, lighting scene modes, and schedule cleaning services. When the edge computing gateway receives remote control commands from mobile terminals, it first verifies the current Nroom status to ensure that high-power electrical devices are not accidentally turned on in unoccupied rooms.

[0080] The following example, taking a standard double room during a complete stay in actual operation, further illustrates the workflow of the interconnected modules of this invention.

[0081] First, the PMS system checks in two guests, registering the number of guests (Ninitial) as 2, and synchronizes this information to the edge computing gateway. The edge computing gateway initializes Nroom to 2, and Nin and Nout are both 0. The smart guest room control unit keeps the power switch active, the air conditioning operates according to the two-person standard, and the fresh air opening is 70%.

[0082] Scenario 1: Guest A leaves, Guest B remains in the room. The time-of-flight sensor array of the porch sensing module detects a person moving outwards. The edge AI algorithm identifies this as a departure event, Nout changes from 0 to 1, and Nroom updates from 2 to 1. The edge computing gateway sends adjustment instructions to the intelligent guest room control host: maintain power supply; fine-tune the air conditioning target temperature from 24 degrees Celsius to 25 degrees Celsius, and reduce the airflow from medium-high speed to medium speed to accommodate the individual's heat load; maintain the current zoning status of the lighting without switching; reduce the fresh air opening from 70% to 60%.

[0083] Scenario 2: Guest B then leaves. The porch sensing module detects another person leaving, Nout changes from 1 to 2, and Nroom updates from 1 to 0. The edge computing gateway starts a five-minute timer. If Nroom remains unchanged within five minutes, the timer expires, and the edge computing gateway sends a deep energy-saving command to the smart room control host: cut off power to all sockets except for the air conditioner's standby power; turn off all room lighting; switch the air conditioner to an energy-saving insulation mode of 28 degrees Celsius in summer or 18 degrees Celsius in winter, with the fan running intermittently at low speed; and close the fresh air valve actuator to the fully closed position. Simultaneously, the edge computing gateway detects that the door lock is locked and the power switch is off, meeting the cleaning criteria, and pushes the room number to the cleaning scheduling system as cleanable.

[0084] Scenario 3: Guest A returns. The porch sensing module detects someone entering, Nin changes from 0 to 1, and Nroom updates from 0 to 1. The edge computing gateway immediately sends a recovery command to the smart guest room control host: close the power switch to restore power to all sockets; prioritize turning on the entrance area lights, then turn on the bedroom main light after a 15-second delay and turn off the entrance light; the air conditioner exits energy-saving mode and returns to 24 degrees Celsius, with the fan speed adjusted to medium; the fresh air intake is increased from fully off to 60%. This process requires no card insertion by the guest.

[0085] Scenario 4: If an unregistered person attempts to enter the room without a card while a guest is out, the door lock sensor reports an unauthorized entry signal. Simultaneously, the porch sensing module detects one person entering, and Nroom changes from 0 to 1. Upon receiving the unauthorized entry signal and the entry event, the edge computing gateway triggers an unauthorized intrusion warning in its local logic judgment unit. The smart guest room control host receives this warning and activates the door lock alarm, while the edge computing gateway pushes the warning to the security mobile terminal, which displays the room number and the time of the incident.

[0086] Scenario 5: If four guests are staying in the double room, i.e., Nroom reaches 4, and this state persists for more than fifteen minutes, the edge computing gateway determines that the number of Nrooms exceeds Ninitial (2), which is equal to the preset yellow warning threshold. It then pushes a yellow over-occupancy warning to the hotel's front desk PMS system to prompt the front desk for verification. If two more guests enter the room ten minutes later, Nroom reaches 6 and the state persists, the edge computing gateway escalates the warning to a red warning and simultaneously pushes it to the security manager's terminal, along with a timeline curve of personnel entry and exit recorded minute by minute over the past two hours for management decision-making.

[0087] Through the collaborative work of the aforementioned modules, this invention achieves accurate occupancy detection in hotel rooms without any image acquisition. Based on the detection results, it enables refined energy-saving control of power supply, air conditioning, lighting, and fresh air systems, as well as intelligent operation and maintenance management for overcrowding, cleaning, and unauthorized intrusion. The edge computing gateway completes all real-time decision-making logic locally, reducing the system's dependence on network bandwidth and improving response speed and reliability. The intervention of the cloud management platform provides group-based chain hotels with the ability to accumulate data and continuously optimize, giving the system good scalability and commercial value.

[0088] The cross-validation mechanism in this invention not only relies on passive infrared sensors and power switches, but can also further integrate door lock status, ultrasonic ranging sensors, and millimeter-wave radar micro-motion detection data within the room. When the porch sensing module fails to report data due to power failure or malfunction, the edge computing gateway can use the linkage status of the passive infrared sensor and power switch as a basis for estimating the number of people in the room. For example, when the power switch is off and the passive infrared sensor has no signal, it is determined that the room is empty; when the power switch is on and the passive infrared sensor is intermittently triggered, it is determined that the room is occupied. However, Nroom does not perform precise numerical correction at this time, but only maintains the binary state of occupied and unoccupied until the porch sensing module resumes operation.

[0089] The people state machine algorithm in this invention also includes an initialization verification sub-step. After check-in is completed in the PMS system, the porch sensing module performs an initial calibration for 30 seconds. During this calibration, it detects whether there are any obstructions or continuous trigger signals in the porch area. If any abnormality is found, the mobile terminal will remind the cleaning staff to clean the porch area to ensure that the sensor's field of vision is unobstructed.

[0090] The time window filtering mechanism in this invention is applicable not only to air conditioning and lighting but also to fresh air control. Within a three-minute window period, the fresh air valve actuator does not react to instantaneous changes in Nroom; it only performs an adjustment once after the window period ends, based on the average or final value of Nroom during the window, thus reducing wear on the actuator motor.

[0091] The cloud management platform in this invention has a policy rollback function. When the edge computing gateway receives new policy parameters from the cloud, if it detects an abnormal increase in energy consumption or an increase in guest complaints during subsequent operation, the edge computing gateway can automatically roll back to the previous stable version of the policy parameters to ensure reliable system operation and guest satisfaction.

[0092] The mobile terminal in this invention also supports guest check-out functionality. When a guest checks out early or leaves temporarily and does not wish for the system to automatically determine this, they can actively submit a check-out command through the mobile terminal. Upon receiving this command, the edge computing gateway will prioritize processing it as a check-out event and decrement Nroom by 1, thereby further enhancing the system's adaptability to special scenarios.

[0093] Example embodiments have been disclosed herein, and while specific terminology has been used, it is for illustrative purposes only and should be construed as such, and is not intended to be limiting. In some instances, it will be apparent to those skilled in the art that features, characteristics, and / or elements described in conjunction with particular embodiments may be used alone, or in combination with features, characteristics, and / or elements described in conjunction with other embodiments, unless otherwise expressly indicated. Therefore, those skilled in the art will understand that various changes in form and detail may be made without departing from the scope of the invention as set forth in the appended claims.

Claims

1. A hotel energy-saving and guest room operation and maintenance method based on human presence perception, characterized in that, include: The direction of people passing through the porch is detected by an array of non-visual sensors installed in the guest room porch area, and the number of people entering and leaving is counted separately. The edge computing gateway receives the number of people entering and leaving, and, combined with the initial number of occupants registered in the PMS system, calculates the real-time number of people in the room using a people state machine algorithm; Based on the real-time number of people in the room, generate and issue multi-strategy control commands; The intelligent guest room control host receives and executes the multi-strategy control instructions to perform energy-saving control of electrical equipment in the guest room or to perform guest room operation and maintenance management; The number of people state machine algorithm includes: The initial room occupancy is set to the initial occupancy number registered in the PMS system; When a person enters the room, the number of people in the room increases by 1; when a person leaves the room, the number of people in the room decreases by 1. The confidence level of the calculated real-time number of people in the room is verified. When the real-time number of people conflicts with the status of the power switch and the data of the PIR auxiliary sensor, the cross-validation mechanism is activated and the correction is made according to the majority consensus principle. The corrected real-time number of people and status changes in the room are pushed to downstream devices.

2. The method according to claim 1, characterized in that, The multi-strategy control instructions include dynamic power access permission management instructions, and the dynamic power access permission management includes: The power switch remains active when the number of people in the room is greater than 0. When the number of people in the room is 0 and continues to exceed the first preset threshold, cut off unnecessary power and keep the air conditioner on standby or switch to energy-saving mode. Power will be automatically restored when the number of people in the room changes from 0 to greater than 0.

3. The method according to claim 1, characterized in that, The multi-strategy control instructions include occupancy adaptive air conditioning control instructions, which include: The target temperature and airflow are dynamically adjusted based on the real-time number of people in the room. When the number of users in real time is 1, the system will operate according to the standard settings. When the number of people in real time is greater than or equal to 2, the target temperature is adjusted according to the average heating and cooling load per person. When the number of people in real time is 0, switch to energy-saving and heat preservation mode.

4. The method according to claim 1, characterized in that, The multi-strategy control instructions include lighting zone control instructions, and the lighting zone control includes: Based on the real-time number of people in the room and the direction of entry detected by the doorway sensing module, the activity area of ​​people is predicted, and the lighting equipment in the corresponding area is turned on or off.

5. The method according to claim 1, characterized in that, The multi-strategy control command includes a fresh air volume dynamic adjustment command, and the fresh air volume dynamic adjustment includes: The opening of the fresh air valve is adjusted in real time according to the number of people in the room. When the number of people in real time is 0, close or control the fresh air valve to operate at the minimum air volume; When the number of people increases in real time, the fresh air volume is increased proportionally.

6. The method according to claim 1, characterized in that, The guest room operation and maintenance management includes overcrowding warning management, which includes: When the number of people in a room exceeds the preset threshold for the number of people registered in the PMS system, and the duration exceeds the second preset threshold, an overcrowding warning message is pushed to the hotel management terminal.

7. The method according to claim 1, characterized in that, The guest room operation and maintenance management includes cleaning demand judgment management, which includes: When the number of people in the room changes from greater than 0 to 0, and the door lock is closed and the power switch is off, it is determined that the guest has left the room, and the room cleaning system is pushed with the cleanable status information; the room usage intensity index is estimated based on the historical data of the number of people in the room, and the cleaning priority is dynamically adjusted based on the usage intensity index.

8. The method according to claim 1, characterized in that, The guest room operation and maintenance management includes illegal check-in detection management, which includes: When the door lock is closed and the door is not opened by swiping a card, and the number of people in the room changes from 0 to greater than 0, an illegal intrusion warning is triggered, and the door lock alarm and security system are activated.

9. A hotel energy-saving and guest room operation and maintenance system based on human presence perception, characterized in that, The system is used to perform the method according to any one of claims 1-8, comprising: The porch sensing module, installed on the top or side frame of the guest room door, uses a non-visual sensor array to detect the direction, speed and body shape characteristics of people passing through the porch, and generates the number of people entering and leaving. An edge computing gateway is communicatively connected to the porch sensing module and has a built-in people state machine algorithm to calculate the real-time number of people in the room based on the number of people entering, leaving, and the initial number of occupants registered in the PMS system, and to execute local policy decisions. The intelligent guest room control host is electrically connected to the electrical equipment in the guest room and is used to receive and execute multi-strategy control commands issued by the edge computing gateway. The cloud management platform communicates with the edge computing gateway via the Internet of Things protocol to receive real-time data, energy consumption data, and abnormal event data from each guest room, and performs big data analysis and strategy optimization. Mobile terminals are provided for use by hotel managers and guests to view room status, receive abnormal alerts, and remotely control guest room equipment.

Images