Robot intelligent control device method and device based on speech recognition

Abstract

The present application provides a voice recognition-based intelligent robot control method and device, which relate to the field of robotics. The method inputs the predicted waiting time for each elevator floor and the set of room numbers to which food items need to be delivered into a pre-trained delivery strategy determination model. A food item delivery strategy is determined that minimizes the total time spent by N delivery robots after all food items to be delivered have been delivered. Based on the food item delivery strategy with the minimum total time spent, food item delivery instructions are sent to each of the N delivery robots. Based on the received food item delivery instructions, the robots sequentially retrieve food items associated with the food item delivery instruction from at least one storage space in a food item temporary storage cabinet, and sequentially take the elevator to deliver the at least one retrieved food item to the door of the guest room corresponding to the at least one room number associated with the food item delivery instruction. This saves labor costs and improves food item delivery efficiency.

Description

Technical Field

[0001] The present application relates to the field of robotics technology, and in particular to a robot intelligent control method and device based on speech recognition. Background Art

[0002] With the development of science and technology and society, intelligent robots have been used in hotels, which can provide delivery services to hotel guests, reduce the workload of hotel staff, and thus improve the hotel's intelligent service level.

[0003] Currently, hotel guests who need food and beverages delivered online can request a robot service through their room phone. The hotel front desk then answers the call, and a staff member places the food and beverages delivered by the delivery driver onto the robot. The front desk attendant then enters the room number and other information, and issues a delivery instruction to the robot, which then performs the requested delivery service.

[0004] However, online food and beverage orders typically arrive at a concentrated time (e.g., lunch takeout typically arrives around noon, and afternoon tea between 3:00 and 4:00 p.m.). Relying on the hotel front desk to coordinate with robots to perform requested delivery services would be labor-intensive, inefficient, and time-consuming, resulting in a poor guest experience. Summary of the Invention

[0005] The present application provides a robot intelligent control method and device based on voice recognition, which is used to solve the problem in the prior art that the labor cost of executing the delivery service requested by the guest is still very high, the efficiency is low, and it takes a long time, resulting in a poor experience for hotel guests.

[0006] In a first aspect, the present application provides a method for intelligent robot control based on speech recognition, which is applied to an intelligent robot control device for speech recognition. The device includes a hotel server, N delivery robots, multiple guest room telephones, and a temporary storage cabinet for food and beverage items. The temporary storage cabinet for food and beverage items includes multiple storage spaces, each storage space has a unique number. Each delivery robot and each guest room telephone are separately connected to the hotel server for communication. The method provided by the present application includes:

[0007] During the target period, the hotel server collects the voice information transmitted from multiple guest room phones to the front desk phone when they call the front desk phone;

[0008] The hotel server determines, based on the voice information of the multiple guest room telephones, a set of identifiers of the guest room telephones of M guest rooms to which the delivery robot is required to deliver food and beverage items, where M and N are both positive integers, and M is greater than N;

[0009] The hotel server determines the guest identity information associated with the room phone number identifier of each room that requires the delivery robot to deliver food and beverage items;

[0010] The hotel server obtains order delivery completion information from the online shopping app account associated with the guest's identity information. The order delivery completion information includes the room number to which the food and beverage items need to be delivered and the storage space number of the food and beverage storage cabinet to be delivered. The room number includes floor information.

[0011] The hotel server inputs the recorded hotel occupancy volume during the target period and the number of guest rooms requiring food and beverage delivery by the delivery robot into a pre-trained first elevator waiting time prediction model to determine a first reference waiting time for the elevator on each floor. The elevator waiting time prediction model is trained by inputting multiple first training samples into a first neural network. Each first training sample includes the historical hotel occupancy volume during the target period, the historical number of guest rooms requiring food and beverage delivery by the delivery robot, and the corresponding historical waiting time for the elevator on each floor.

[0012] The hotel server inputs the actual waiting time of the elevator on each floor in the two previous historical time periods equal to the duration of the target time period, and the actual waiting time of the elevator on each floor in the previous historical time period equal to and continuous with the duration of the target time period, into a pre-trained second elevator waiting time prediction model to determine a second reference waiting time for the elevator on each floor, wherein the second elevator waiting time prediction model is trained by inputting multiple second training samples into a long short-term memory model, each second training sample including the historical waiting time of the elevator on each floor in the two previous historical time periods equal to the duration of the historical target time period, the historical waiting time of the elevator on each floor in the previous historical time period equal to and continuous with the duration of the historical target time period, and the historical waiting time of the elevator on each floor in the historical target time period;

[0013] The hotel server determines the predicted waiting time for the elevator on each floor based on the first reference waiting time for the elevator on each floor and the second reference waiting time for the elevator on each floor;

[0014] The hotel server inputs the predicted waiting time for elevators on each floor and the set of room numbers to which food items need to be delivered into a pre-trained delivery strategy determination model, and determines the food delivery strategy that minimizes the total time spent by N delivery robots after all food items to be delivered are delivered. The delivery strategy determination model is trained by inputting multiple third training samples into the second neural network. Each third training sample includes the historical predicted waiting time for elevators on each floor, the historical set of room numbers to which food items need to be delivered, and the corresponding historical food delivery strategy that minimizes the total time spent after all food items to be delivered are delivered.

[0015] The hotel server sends a food item delivery instruction to each of the N delivery robots based on the food item delivery strategy that minimizes the total delivery time. Each food item delivery instruction includes the number of at least one storage space of the food item to be delivered in the food item temporary storage cabinet and the number of at least one room to be delivered.

[0016] Based on the received food item delivery instructions, N delivery robots take out food items associated with the food item delivery instructions from at least one storage space of the food item temporary storage cabinet in turn, and take the elevator in turn to deliver the at least one food item taken out to the door of the guest room corresponding to at least one room number associated with the food item delivery instruction.

[0017] In one possible implementation, the hotel server determines, based on voice information from multiple guest room telephones, a set of identifiers of M guest room telephones to which a delivery robot is required to deliver food and beverage items, including:

[0018] The hotel server converts the voice messages of multiple guest room phones into corresponding text messages;

[0019] The hotel server compares each text message with each standard text message in a preset standard text library for requesting food and beverage delivery.

[0020] When the similarity between the text message and any standard text message is greater than a set similarity threshold, the hotel server determines that the guest room corresponding to the text message and the guest room phone number thereof need a delivery robot to deliver food and beverage items;

[0021] The hotel server aggregates the identifiers of the telephone numbers of multiple guest rooms that require the delivery robot to deliver food and beverage items, and obtains an identifier set of the telephone numbers of M guest rooms that require the delivery robot to deliver food and beverage items.

[0022] In one possible implementation, the hotel server determines, based on voice information from multiple guest room telephones, a set of identifiers of M guest room telephones to which a delivery robot is required to deliver food and beverage items, including:

[0023] The hotel server inputs the voice information of the multiple guest room telephones into a pre-set semantic recognition model to respectively identify the semantics corresponding to the voice information of the multiple guest room telephones;

[0024] When the semantics corresponding to the voice information of any guest room phone indicate that a delivery robot is required to deliver food and beverage items, the hotel server determines that the guest room corresponding to the voice information is located in a room that requires a delivery robot to deliver food and beverage items;

[0025] The hotel server aggregates the identifiers of the telephone numbers of multiple guest rooms that require the delivery robot to deliver food and beverage items, and obtains an identifier set of the telephone numbers of M guest rooms that require the delivery robot to deliver food and beverage items.

[0026] In one possible implementation, the hotel server determines the predicted waiting time for the elevator on each floor based on the first reference waiting time for the elevator on each floor and the second reference waiting time for the elevator on each floor, including:

[0027] The hotel server is based on the formula T 11 =K1T1+K2T2, determine the predicted waiting time for the elevator on each floor, where T 11 is the predicted waiting time for the elevator on any floor, T1 is the first reference waiting time for the elevator on any floor, T2 is the second reference waiting time for the elevator on any floor, K1 is the first weighting coefficient, and K2 is the second weighting coefficient.

[0028] In one possible implementation, the delivery robot includes a manipulator and a placement tray. Based on a received food item delivery instruction, N delivery robots sequentially retrieve food items associated with the food item delivery instruction from at least one storage space of a food item temporary storage cabinet, and sequentially take an elevator to sequentially deliver the retrieved at least one food item to a guest room door corresponding to at least one room number associated with the food item delivery instruction, including:

[0029] For any one of the N delivery robots, receiving a food item delivery instruction;

[0030] Identifying the number of at least one storage space of a food item temporary storage cabinet and at least one room number to be delivered in the food item delivery instruction;

[0031] Move to a side of the storage space corresponding to the number of at least one storage space;

[0032] Extending a robotic arm to grab a food item to be delivered that is located in at least one storage space corresponding to the number, and placing the at least one food item to be delivered on a placement tray;

[0033] Generate a navigation route based on the address of the food storage cabinet and the address associated with at least one room number to be delivered;

[0034] Based on the navigation route, take the elevator with the food and beverage items to be delivered and move to the door of at least one guest room associated with the room number to be delivered.

[0035] In a second aspect, the present application also provides a robot intelligent control device based on voice recognition, comprising a hotel server, N delivery robots, multiple guest room telephones, and a temporary storage cabinet for catering items. The temporary storage cabinet for catering items comprises multiple storage spaces, each storage space comprises a unique number, and each delivery robot and each guest room telephone are respectively connected to the hotel server for communication, wherein,

[0036] The hotel server is used to collect the voice information transmitted from multiple guest room phones to the front desk phone during the target period.

[0037] The hotel server is further configured to determine, based on the voice information of the plurality of guest room telephones, a set of identifiers of the guest room telephones of M guest rooms to which the delivery robot is required to deliver food and beverage items, where M and N are both positive integers and M is greater than N;

[0038] The hotel server is further configured to determine the guest identity information associated with the room phone number identifier of each room to which the delivery robot is required to deliver food and beverage items;

[0039] The hotel server is also used to obtain order delivery completion information from the online shopping application account associated with the guest's identity information. The order delivery completion information includes the room number to which the food and beverage items need to be delivered and the storage space number of the food and beverage temporary storage cabinet to be delivered. The room number includes floor information.

[0040] The hotel server is further configured to input the recorded hotel occupancy count within the target time period and the number of guest rooms requiring food and beverage delivery by the delivery robot into a pre-trained first elevator waiting time prediction model to determine a first reference waiting time for the elevator on each floor, wherein the elevator waiting time prediction model is trained by inputting a plurality of first training samples into a first neural network, each first training sample including a historical hotel occupancy count within the target time period, a historical number of guest rooms requiring food and beverage delivery by the delivery robot, and a corresponding historical waiting time for the elevator on each floor;

[0041] The hotel server is further configured to input the actual waiting time of the elevator on each floor in the two previous historical time periods equal to the duration of the target time period, and the actual waiting time of the elevator on each floor in the previous historical time period equal to and continuous with the duration of the target time period, into a pre-trained second elevator waiting time prediction model to determine a second reference waiting time for the elevator on each floor, wherein the second elevator waiting time prediction model is trained by inputting a plurality of second training samples into a long short-term memory model, each second training sample including the historical waiting time of the elevator on each floor in the two previous historical time periods equal to the duration of the historical target time period, the historical waiting time of the elevator on each floor in the previous historical time period equal to and continuous with the duration of the historical target time period, and the historical waiting time of the elevator on each floor in the historical target time period;

[0042] The hotel server is further configured to determine a predicted waiting time for the elevator on each floor based on the first reference waiting time for the elevator on each floor and the second reference waiting time for the elevator on each floor;

[0043] The hotel server is further configured to input the predicted waiting time for the elevator on each floor and the set of room numbers to which food items need to be delivered into a pre-trained delivery strategy determination model, and determine a food item delivery strategy that minimizes the total time spent by the N delivery robots after all food items to be delivered are delivered. The delivery strategy determination model is trained by inputting multiple third training samples into the second neural network, each third training sample including the historical predicted waiting time for the elevator on each floor, the historical set of room numbers to which food items need to be delivered, and the corresponding historical food item delivery strategy that minimizes the total time spent after all food items to be delivered are delivered.

[0044] The hotel server is further configured to send food item delivery instructions to each of the N delivery robots based on the food item delivery strategy that minimizes the total delivery time, wherein each food item delivery instruction includes the number of at least one storage space of the food item to be delivered in the food item temporary storage cabinet and the number of at least one room to be delivered;

[0045] The N delivery robots are respectively configured to, based on a received food item delivery instruction, sequentially retrieve food items associated with the food item delivery instruction from at least one storage space of the food item temporary storage cabinet;

[0046] The N delivery robots are further configured to take the elevator in turn to deliver the taken-out at least one food and beverage item to the guest room door a corresponding to at least one room number associated with the food and beverage item delivery instruction.

[0047] In one possible implementation, a hotel server is specifically configured to convert voice information from multiple guest room telephones into corresponding text information; for each text message, the text message is compared for similarity with each standard text message in a preset standard text library requesting the delivery of food and beverage items; when the similarity between the text message and any standard text message is greater than a set similarity threshold, it is determined that the guest room to which the guest room telephone corresponding to the text message is located requires a delivery robot to deliver food and beverage items; the identifiers of multiple guest room telephones that require a delivery robot to deliver food and beverage items are aggregated to obtain a set of identifiers of the guest room telephones of M guest rooms that require a delivery robot to deliver food and beverage items.

[0048] In one possible implementation, a hotel server is specifically configured to input voice information from multiple guest room telephones into a preset semantic recognition model to respectively identify the semantics corresponding to the voice information from the multiple guest room telephones; when the semantics corresponding to the voice information from any guest room telephone represent the need for a delivery robot to deliver food and beverage items, determine that the guest room where the guest room telephone corresponding to the voice information is located needs a delivery robot to deliver food and beverage items; aggregate the identifiers of multiple guest room telephones that need a delivery robot to deliver food and beverage items, and obtain a set of identifiers of the guest room telephones of M guest rooms that need a delivery robot to deliver food and beverage items.

[0049] In a possible implementation, the hotel server is specifically configured to calculate the 11 =K1T1+K2T2, determine the predicted waiting time for the elevator on each floor, where T 11 is the predicted waiting time for the elevator on any floor, T1 is the first reference waiting time for the elevator on any floor, T2 is the second reference waiting time for the elevator on any floor, K1 is the first weighting coefficient, and K2 is the second weighting coefficient.

[0050] In one possible embodiment, each delivery robot includes a manipulator and a placement tray, and each delivery robot among the N delivery robots is used to receive a food item delivery instruction; identify the number of at least one storage space of a food item temporary storage cabinet and at least one room number to be delivered in the food item delivery instruction; move to one side of the storage space corresponding to the number of at least one storage space in turn; extend the manipulator to grab the food item to be delivered in the storage space corresponding to the at least one number, and place the at least one food item to be delivered on the placement tray; generate a navigation route based on the address of the food item temporary storage cabinet and the address associated with at least one room number to be delivered; based on the navigation route, carry the food item to be delivered and take the elevator to the door of the guest room associated with at least one room number to be delivered in turn.

[0051] The present application provides a robot intelligent control method and device based on voice recognition. The hotel server determines the identification set of the guest room telephones of M guest rooms that require the delivery robot to deliver food and beverage items based on the voice information of multiple guest room telephones.

[0052] The hotel server determines the guest identity information associated with the room phone number identifier for each guest room to which a food delivery robot is requested. As can be appreciated, guests checking into a hotel are required to register using their ID cards, so the hotel server records the correspondence between the guest identity information and the room number in which the guest is staying. Since room numbers correspond to the room phone identifiers within the guest rooms, the hotel server can determine the guest identity information associated with the room phone identifiers.

[0053] In this way, the hotel server can obtain the order delivery completion information under the online shopping application account associated with the guest's identity information. The order delivery completion information includes the room number to which the food and beverage items need to be delivered and the storage space number of the food and beverage item temporary storage cabinet to be delivered. The room number carries the floor information.

[0054] The hotel server inputs the recorded hotel occupancy during the target time period and the number of guest rooms requiring food and beverage delivery by the delivery robot into a pre-trained first elevator wait time prediction model to determine a first reference wait time for each elevator floor. Because the elevator wait time prediction model is trained by inputting multiple first training samples into the first neural network, each first training sample includes the historical hotel occupancy during the target time period, the historical number of guest rooms requiring food and beverage delivery by the delivery robot, and the corresponding historical wait time for each elevator floor, this allows for a high degree of accuracy in determining the first reference wait time for each elevator floor.

[0055] In addition, the hotel server also inputs the actual waiting time for each elevator floor in the two previous historical time periods equal to the duration of the target time period, and the actual waiting time for each elevator floor in the previous historical time period equal to and continuous with the duration of the target time period, into the pre-trained second elevator waiting time prediction model to determine the second reference waiting time for each elevator floor. Because the second elevator waiting time prediction model is trained by inputting multiple second training samples into the long short-term memory model, each second training sample includes the historical waiting time for each elevator floor in the two previous historical time periods equal to the duration of the historical target time period, the historical waiting time for each elevator floor in the previous historical time period equal to and continuous with the duration of the historical target time period, and the historical waiting time for each elevator floor in the historical target time period. In this way, the second reference waiting time for each elevator floor is determined with high accuracy.

[0056] The hotel server determines a predicted elevator wait time for each floor based on the first reference wait time for each floor and the second reference wait time for each floor. Because the first reference wait time is highly accurate, the second reference wait time is also highly accurate. Therefore, the predicted elevator wait time for each floor based on the first and second reference wait times is also highly accurate.

[0057] The hotel server inputs the predicted waiting time for each elevator floor and the set of room numbers to which food items need to be delivered into a pre-trained delivery strategy determination model. This determines the food delivery strategy that minimizes the total time it takes for N delivery robots to deliver all the food items to be delivered. Since the delivery strategy determination model is trained by inputting multiple third training samples into the second neural network, each third training sample includes the historically predicted waiting time for each elevator floor, the historical set of room numbers to which food items need to be delivered, and the corresponding historical food delivery strategy that minimized the total time it takes for all the food items to be delivered. This determines the food delivery strategy that minimizes the total time it takes for all the food items to be delivered, resulting in exceptional efficiency. Furthermore, because the time it takes into account elevator wait times, it is also highly reliable.

[0058] Based on a food item delivery strategy that minimizes total delivery time, the hotel server sends food item delivery instructions to each of the N delivery robots. Each food item delivery instruction includes the number of at least one storage space in the food item temporary storage cabinet and the number of at least one room to be delivered. Based on the received food item delivery instructions, the N delivery robots sequentially remove the food item associated with the food item delivery instruction from the at least one storage space in the food item temporary storage cabinet, and then take the elevator to deliver the at least one food item to the door of the guest room corresponding to the at least one room number associated with the food item delivery instruction. This allows the food item to be delivered without any human intervention, saving labor costs and further improving food item delivery efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, a brief introduction will be given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0060] Figure 1 A schematic diagram of the architecture of a robot intelligent control device based on speech recognition provided in an embodiment of the present application;

[0061] Figure 2 A flowchart of a robot intelligent control method based on speech recognition provided in an embodiment of the present application. DETAILED DESCRIPTION

[0062] To make the purpose, technical solutions, and advantages of the embodiments of the present application more clear, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments made by ordinary technicians in this field based on the inspiration of these embodiments fall within the scope of protection of this application.

[0063] The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequential order. It should be understood that the numbers used in this way are interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0064] See also Figure 1 The embodiment of the present application provides a robot intelligent control method based on speech recognition, which is applied to a robot intelligent control device based on speech recognition. The device includes a hotel server 100, N delivery robots 103, multiple guest room telephones, and a temporary storage cabinet 102 for food and beverage items. The temporary storage cabinet 102 for food and beverage items includes multiple storage spaces, each storage space includes a unique number, and each delivery robot 103 and each guest room telephone are respectively connected to the hotel server 100 for communication. Figure 2 As shown, the method provided in the embodiment of the present application includes:

[0065] S201: The hotel server 100 collects voice information transmitted from multiple guest room phones to the front desk phone when the guest room phones make calls to the front desk phone during a target period.

[0066] The target time period may be a time period for concentrated food delivery, for example, 11:30 AM to 1:00 PM. The voice message may include, but is not limited to, "Please pick up my takeaway," "Hello, please pick up my food," or "Send me my meal."

[0067] S202: The hotel server 100 determines, based on the voice information of the multiple guest room telephones, a set of identifications of the guest room telephones of the M guest rooms 104 to which the delivery robot 103 is required to deliver food and beverage items.

[0068] Wherein, M and N are both positive integers, and M is greater than N. Typically, hotels use a relatively small number of delivery robots 103 for cost considerations. For example, the number N of delivery robots 103 is generally less than 5. However, the delivery robots 103 typically deliver a large number of food and beverage items, typically greater than 20. Therefore, a reasonable food and beverage item delivery strategy needs to be determined to minimize the total time spent.

[0069] For example, specific implementations of S202 may include but are not limited to the following two methods:

[0070] The first one:

[0071] Step 1: The hotel server 100 converts the voice messages of multiple guest room telephones into corresponding text messages.

[0072] Step 2: The hotel server 100 compares each text message with each standard text message in a preset standard text library for requesting food and beverage delivery for similarity.

[0073] Step 3: When the similarity between the text message and any standard text message is greater than a set similarity threshold, the hotel server 100 determines that the guest room 104 where the guest room phone number corresponding to the text message is located needs the delivery robot 103 to deliver food and beverage items.

[0074] Step 4: The hotel server 100 aggregates the identifiers of the telephone numbers of multiple guest rooms that require the delivery robot 103 to deliver food and beverage items, and obtains an identifier set of the telephone numbers of the M guest rooms 104 that require the delivery robot 103 to deliver food and beverage items.

[0075] The second type:

[0076] Step A: The hotel server 100 inputs the voice information of the multiple guest room telephones into a preset semantic recognition model to respectively recognize the semantics corresponding to the voice information of the multiple guest room telephones.

[0077] Step B: When the semantics corresponding to the voice information of any guest room phone indicate that the delivery robot 103 is required to deliver food and beverage items, the hotel server 100 determines that the guest room 104 where the guest room phone corresponding to the voice information is located needs the delivery robot 103 to deliver food and beverage items.

[0078] Step C: The hotel server 100 aggregates the identifiers of the telephone numbers of multiple guest rooms that require the delivery robot 103 to deliver food and beverage items, and obtains an identifier set of the telephone numbers of the M guest rooms 104 that require the delivery robot 103 to deliver food and beverage items.

[0079] S203 : The hotel server 100 determines the guest identity information associated with the room phone identifier of each room 104 to which the delivery robot 103 is required to deliver food and beverage items.

[0080] S204: The hotel server 100 obtains order delivery completion information from the online shopping application account associated with the guest's identity information.

[0081] The order delivery completion information includes the room number to which the food and beverage items need to be delivered and the number of the storage space of the food and beverage items to be delivered in the food and beverage item temporary storage cabinet 102 , and the room number carries floor information.

[0082] S205: The hotel server 100 inputs the recorded hotel occupancy volume during the target period and the number of guest rooms 104 that require the delivery robot 103 to deliver food and beverage items into the pre-trained first elevator 103 waiting time prediction model to determine the first reference waiting time for the elevator 103 on each floor.

[0083] Among them, the elevator 103 waiting time prediction model is obtained by inputting multiple first training samples into the first neural network for training. Each first training sample includes the hotel occupancy volume during the historical target period, the number of guest rooms 104 that historically required the delivery robot 103 to deliver food and beverage items, and the corresponding historical waiting time of the elevator 103 on each floor.

[0084] S206: The hotel server 100 inputs the actual waiting time of the elevator 103 on each floor in the first two historical time periods equal to the length of the target time period and the actual waiting time of the elevator 103 on each floor in the first historical time period equal to and continuous with the length of the target time period into the pre-trained second elevator 103 waiting time prediction model to determine the second reference waiting time of the elevator 103 on each floor.

[0085] Among them, the second elevator 103 waiting time prediction model is obtained by inputting multiple second training samples into the long short-term memory model for training, each second training sample includes the historical waiting time of the elevator 103 on each floor in the first two historical time periods equal to the length of the historical target time period, the historical waiting time of the elevator 103 on each floor in the first historical time period equal to and continuous with the length of the historical target time period, and the historical waiting time of the elevator 103 on each floor in the historical target time period.

[0086] S207: The hotel server 100 determines the predicted waiting time of the elevator 103 on each floor according to the first reference waiting time of the elevator 103 on each floor and the second reference waiting time of the elevator 103 on each floor.

[0087] For example, the hotel server 100 can calculate the value of the formula T 11 =K1T1+K2T2, determine the predicted waiting time of the elevator 103 on each floor, where T 11 is the predicted waiting time of the elevator 103 on any floor, T1 is the first reference waiting time of the elevator 103 on any floor, T2 is the second reference waiting time of the elevator 103 on any floor, K1 is the first weighting coefficient, and K2 is the second weighting coefficient. For example, K1 can be equal to 0.5, and K2 can be equal to 0.5.

[0088] S208: The hotel server 100 inputs the predicted waiting time of the elevator 103 on each floor and the set of room numbers to which the food and beverage items need to be delivered into the pre-trained delivery strategy determination model, and determines the food and beverage item delivery strategy with the shortest total time after N delivery robots 103 have delivered all the food and beverage items to be delivered.

[0089] Among them, the distribution strategy determination model is obtained by inputting multiple third training samples into the second neural network for training. Each third training sample includes the predicted waiting time of the elevator 103 on each floor in history, the set of room numbers to which catering items need to be delivered in history, and the corresponding catering item delivery strategy with the shortest total time after all catering items to be delivered have been delivered in history.

[0090] For example, the food delivery strategy with the shortest total time includes but is not limited to assigning the delivery room numbers of the same floor to the same delivery robot 103, assigning the delivery room numbers of adjacent floors to the same delivery robot 103, etc., which are not limited here.

[0091] S209: The hotel server 100 sends food and beverage delivery instructions to the N delivery robots 103 based on the food and beverage delivery strategy with the shortest total time.

[0092] Each food item delivery instruction includes the number of at least one storage space of the food item to be delivered in the food item temporary storage cabinet 102 and at least one room number to be delivered.

[0093] S210: N delivery robots 103 take out the food items associated with the food item delivery instruction from at least one storage space of the food item temporary storage cabinet 102 in turn based on the received food item delivery instruction, and take the elevator 103 in turn to deliver the at least one food item taken out to the door of the guest room 104 corresponding to at least one room number associated with the food item delivery instruction.

[0094] Exemplarily, the delivery robot 103 includes a manipulator and a placement tray, and S202 can be specifically implemented as follows:

[0095] For any one of the N delivery robots 103 , a food item delivery instruction is received.

[0096] Identify the number of at least one storage space of the food item temporary storage cabinet 102 and at least one room number to be delivered in the food item delivery instruction.

[0097] Move in sequence to one side of the storage space corresponding to the number of at least one storage space of the food and beverage item temporary storage cabinet 102.

[0098] The robotic arm is extended to grab the food and beverage item to be delivered in at least one storage space corresponding to the number, and the at least one food and beverage item to be delivered is placed on the placement tray.

[0099] For example, if at least one storage space is numbered "001," "002," and "003," the robot moves to the side of the storage space numbered "001" of the food item temporary storage cabinet 102, extends the robot arm, and places the food items in the storage space numbered "001" on the placement tray. Then, the robot moves to the side of the storage space numbered "002" of the food item temporary storage cabinet 102, extends the robot arm, and places the food items in the storage space numbered "002" on the placement tray. Finally, the robot moves to the side of the storage space numbered "003" of the food item temporary storage cabinet 102, extends the robot arm, and places the food items in the storage space numbered "003" on the placement tray.

[0100] A navigation route is generated based on the address of the food item storage cabinet 102 and the address associated with at least one room number to be delivered. Based on the navigation route, the food items to be delivered are taken in turn by elevator 103 to the door of the guest room 104 associated with at least one room number to be delivered.

[0101] Still see Figure 1 , the embodiment of the present application also provides a robot intelligent control device based on voice recognition. It should be noted that the basic principle and technical effects of the robot intelligent control device based on voice recognition provided by the embodiment of the present invention are the same as those of the above-mentioned embodiment. For the sake of brief description, for parts not mentioned in the embodiment of the present invention, please refer to the corresponding content in the above-mentioned embodiment. The device includes a hotel server 100, N delivery robots 103, multiple guest room telephones and a catering item temporary storage cabinet 102. The catering item temporary storage cabinet 102 includes multiple storage spaces, each storage space includes a unique number, each delivery robot 103 and each guest room telephone are respectively communicated with the hotel server 100, wherein,

[0102] The hotel server 100 is used to collect voice information transmitted from multiple guest room phones to the front desk phone when making calls to the front desk phone during a target period;

[0103] The hotel server 100 is further configured to determine, based on the voice information of the plurality of guest room telephones, a set of identifiers of the guest room telephones of M guest rooms 104 to which the delivery robot 103 is required to deliver food and beverage items, where M and N are both positive integers, and M is greater than N;

[0104] The hotel server 100 is further configured to determine the guest identity information associated with the room phone number identifier of each room 104 to which the delivery robot 103 is required to deliver food and beverage items;

[0105] The hotel server 100 is further configured to obtain order delivery completion information from the online shopping application account associated with the guest's identity information, wherein the order delivery completion information includes the room number to which the food and beverage item needs to be delivered and the storage space number of the food and beverage item temporary storage cabinet 102 to be delivered, wherein the room number includes floor information;

[0106] The hotel server 100 is further configured to input the recorded hotel occupancy count within the target time period and the number of guest rooms 104 requiring food and beverage delivery by the delivery robot 103 into a pre-trained first elevator 103 waiting time prediction model to determine a first reference waiting time for the elevator 103 on each floor. The elevator 103 waiting time prediction model is trained by inputting multiple first training samples into a first neural network, each first training sample including the historical hotel occupancy count within the target time period, the historical number of guest rooms 104 requiring food and beverage delivery by the delivery robot 103, and the corresponding historical waiting time for the elevator 103 on each floor.

[0107] The hotel server 100 is further configured to input the actual waiting time of the elevator 103 on each floor in the two previous historical time periods equal to the duration of the target time period, and the actual waiting time of the elevator 103 on each floor in the previous historical time period equal to and continuous with the duration of the target time period, into a pre-trained second elevator 103 waiting time prediction model to determine a second reference waiting time for the elevator 103 on each floor, wherein the second elevator 103 waiting time prediction model is trained by inputting multiple second training samples into a long short-term memory model, each second training sample including the historical waiting time of the elevator 103 on each floor in the two previous historical time periods equal to the duration of the historical target time period, the historical waiting time of the elevator 103 on each floor in the previous historical time period equal to and continuous with the duration of the historical target time period, and the historical waiting time of the elevator 103 on each floor in the historical target time period;

[0108] The hotel server 100 is further configured to determine a predicted waiting time for the elevator 103 on each floor based on the first reference waiting time for the elevator 103 on each floor and the second reference waiting time for the elevator 103 on each floor;

[0109] The hotel server 100 is further configured to input the predicted waiting time for the elevator 103 on each floor and the set of room numbers to which food items need to be delivered into a pre-trained delivery strategy determination model, and determine a food item delivery strategy that minimizes the total time spent after all food items to be delivered are delivered by the N delivery robots 103. The delivery strategy determination model is trained by inputting multiple third training samples into the second neural network, each third training sample including the historical predicted waiting time for the elevator 103 on each floor, the historical set of room numbers to which food items need to be delivered, and the corresponding historical food item delivery strategy that minimizes the total time spent after all food items to be delivered are delivered.

[0110] The hotel server 100 is further configured to send food item delivery instructions to each of the N delivery robots 103 based on the food item delivery strategy with the shortest total delivery time, wherein each food item delivery instruction includes the number of at least one storage space of the food item to be delivered in the food item temporary storage cabinet 102 and the number of at least one room to be delivered;

[0111] The N delivery robots 103 are respectively configured to, based on the received food item delivery instructions, sequentially take out food items associated with the food item delivery instructions from at least one storage space of the food item temporary storage cabinet 102;

[0112] The N delivery robots 103 are further configured to take the elevator 103 in sequence and deliver the taken-out at least one food and beverage item to the door of the guest room 104 corresponding to at least one room number associated with the food and beverage item delivery instruction.

[0113] In one possible embodiment, the hotel server 100 is specifically used to convert the voice information of multiple guest room telephones into corresponding text information; for each text message, the text information is compared with each standard text message in a preset standard text library for requesting the delivery of food and beverage items; when the similarity between the text information and any standard text information is greater than a set similarity threshold, it is determined that the guest room 104 where the guest room telephone corresponding to the text information is located needs the delivery robot 103 to deliver food and beverage items; the identifiers of multiple guest room telephones that need the delivery robot 103 to deliver food and beverage items are aggregated to obtain a set of identifiers of the guest room telephones of M guest rooms 104 that need the delivery robot 103 to deliver food and beverage items.

[0114] In one possible embodiment, the hotel server 100 is specifically used to input the voice information of multiple guest room telephones into a preset semantic recognition model to respectively identify the semantics corresponding to the voice information of the multiple guest room telephones; when the semantics corresponding to the voice information of any guest room telephone represents the need for the delivery robot 103 to deliver food and beverage items, it is determined that the guest room 104 where the guest room telephone corresponding to the voice information is located needs the delivery robot 103 to deliver food and beverage items; the identifiers of multiple guest room telephones that need the delivery robot 103 to deliver food and beverage items are aggregated to obtain a set of identifiers of the guest room telephones of M guest rooms 104 that need the delivery robot 103 to deliver food and beverage items.

[0115] In a possible implementation, the hotel server 100 is specifically configured to calculate the 11 =K1T1+K2T2, determine the predicted waiting time of the elevator 103 on each floor, where T 11 is the predicted waiting time of the elevator 103 on any floor, T1 is the first reference waiting time of the elevator 103 on any floor, T2 is the second reference waiting time of the elevator 103 on any floor, K1 is the first weighting coefficient, and K2 is the second weighting coefficient.

[0116] In one possible embodiment, each delivery robot 103 includes a manipulator and a placement plate, and each delivery robot 103 among the N delivery robots 103 is used to receive a food item delivery instruction; identify the number of at least one storage space of the food item temporary storage cabinet 102 and at least one room number to be delivered in the food item delivery instruction; move to one side of the storage space corresponding to the number of at least one storage space; extend the manipulator to grab the food item to be delivered in the storage space corresponding to the at least one number, and place the at least one food item to be delivered on the placement plate; generate a navigation route based on the address of the food item temporary storage cabinet 102 and the address associated with at least one room number to be delivered; based on the navigation route, carry the food items to be delivered and take the elevator 103 in turn to move to the door of the guest room 104 associated with at least one room number to be delivered.

[0117] The present application provides a robot intelligent control method based on voice recognition. The hotel server 100 determines the identification set of the guest room telephones of M guest rooms 104 to which the delivery robot 103 delivers food and beverage items based on the voice information of multiple guest room telephones.

[0118] The hotel server 100 determines the guest identity information associated with the room phone number identifier of each guest room 104 to which the delivery robot 103 is to deliver food and beverage items. As can be appreciated, guests checking into a hotel are required to register using their ID cards, so the hotel server 100 records the correspondence between the guest identity information and the room number in which the guest is staying. Since the room number corresponds to the room phone number identifier in each guest room 104, the hotel server 100 can determine the guest identity information associated with the room phone number identifier.

[0119] In this way, the hotel server 100 can obtain the order delivery completion information under the online shopping application account associated with the guest identity information, wherein the order delivery completion information includes the room number to which the food and beverage items need to be delivered and the storage space number of the food and beverage item temporary storage cabinet 102 to be delivered, and the room number carries the floor information.

[0120] The hotel server 100 inputs the recorded hotel occupancy during the target time period and the number of guest rooms 104 requiring food and beverage delivery by the delivery robot 103 into a pre-trained first elevator 103 wait time prediction model to determine a first reference wait time for each floor's elevator 103. Because the elevator 103 wait time prediction model is trained by inputting multiple first training samples into the first neural network, each first training sample includes the historical hotel occupancy during the target time period, the historical number of guest rooms 104 requiring food and beverage delivery by the delivery robot 103, and the corresponding historical wait time for each floor's elevator 103. This allows for a high degree of accuracy in determining the first reference wait time for each floor's elevator 103.

[0121] In addition, the hotel server 100 also inputs the actual waiting time of each elevator 103 on each floor during the two previous historical time periods equal to the target time period, and the actual waiting time of each elevator 103 on each floor during the previous historical time period equal to and consecutive to the target time period, into a pre-trained second elevator 103 waiting time prediction model to determine a second reference waiting time for each elevator 103 on each floor. Because the second elevator 103 waiting time prediction model is trained by inputting multiple second training samples into the long short-term memory model, each second training sample includes the historical waiting time of each elevator 103 on each floor during the two previous historical time periods equal to the target time period, the historical waiting time of each elevator 103 on each floor during the previous historical time period equal to and consecutive to the target time period, and the historical waiting time of each elevator 103 on each floor during the target time period. In this way, the second reference waiting time for each elevator 103 on each floor is determined with high accuracy.

[0122] The hotel server 100 determines the predicted waiting time for the elevator 103 on each floor based on the first reference waiting time for each elevator 103 and the second reference waiting time for each elevator 103 on each floor. Since the first reference waiting time is highly accurate and the second reference waiting time is also highly accurate, the predicted waiting time for the elevator 103 on each floor based on the first reference waiting time and the second reference waiting time is also highly accurate.

[0123] The hotel server 100 inputs the predicted waiting time for each elevator 103 on each floor and the set of room numbers to which food items need to be delivered into a pre-trained delivery strategy determination model, and determines the food item delivery strategy that minimizes the total time spent by the N delivery robots 103 after all food items to be delivered have been delivered. Because the delivery strategy determination model is trained by inputting multiple third training samples into the second neural network, each third training sample includes the historically predicted waiting time for each elevator 103 on each floor, the historical set of room numbers to which food items need to be delivered, and the corresponding historical food item delivery strategy that minimized the total time spent after all food items to be delivered have been delivered. In this way, the determined food item delivery strategy minimizes the total time spent after all food items to be delivered have been delivered, resulting in particularly high efficiency.

[0124] Based on the food item delivery strategy that minimizes total delivery time, the hotel server 100 sends food item delivery instructions to each of the N delivery robots 103. Each food item delivery instruction includes the number of the food item to be delivered in at least one storage space of the food item temporary storage cabinet 102 and the number of at least one room to be delivered. Based on the received food item delivery instructions, the N delivery robots 103 sequentially remove the food item associated with the food item delivery instruction from the at least one storage space of the food item temporary storage cabinet 102, and then take the elevator 103 in turn to deliver the at least one food item to the door of the guest room 104 corresponding to the at least one room number associated with the food item delivery instruction. In this way, the food item delivery described above does not require any human intervention, saving labor costs and further improving food item delivery efficiency.

[0125] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A robot intelligent control method based on speech recognition, characterized in that: A robot intelligent control device for speech recognition includes a hotel server, N delivery robots, multiple guest room telephones, and a temporary storage cabinet for food and beverage items. The temporary storage cabinet includes multiple storage spaces, each storage space includes a unique number. Each delivery robot and each guest room telephone are respectively connected to the hotel server for communication. The method includes: The hotel server collects voice information transmitted by the plurality of guest room phones to the front desk phone when the guest room phones make calls to the front desk phone during a target period; The hotel server determines, based on the voice information of the plurality of guest room telephones, a set of identifiers of guest room telephones of M guest rooms to which the delivery robot is required to deliver food and beverage items, where M and N are both positive integers and M is greater than N; The hotel server determines the guest identity information associated with the room phone identifier of each room to which the delivery robot is required to deliver food and beverage items; The hotel server obtains order delivery completion information from the online shopping application account associated with the guest identity information, wherein the order delivery completion information includes the room number to which the food and beverage item needs to be delivered and the number of the storage space of the food and beverage item temporary storage cabinet to be delivered, and the room number carries floor information; The hotel server inputs the recorded hotel occupancy volume during the target time period and the number of guest rooms requiring food and beverage delivery by the delivery robot into a pre-trained first elevator waiting time prediction model to determine a first reference waiting time for the elevator on each floor, wherein the elevator waiting time prediction model is trained by inputting multiple first training samples into a first neural network, each of the first training samples including the historical hotel occupancy volume during the target time period, the historical number of guest rooms requiring food and beverage delivery by the delivery robot, and the corresponding historical waiting time for the elevator on each floor; The hotel server inputs the actual waiting time of the elevator on each floor in the first two historical time periods equal to the duration of the target time period, and the actual waiting time of the elevator on each floor in the first historical time period equal to and continuous with the duration of the target time period, into a pre-trained second elevator waiting time prediction model to determine a second reference waiting time for the elevator on each floor, wherein the second elevator waiting time prediction model is trained by inputting multiple second training samples into a long short-term memory model, each of the second training samples including the historical waiting time of the elevator on each floor in the first two historical time periods equal to the duration of the historical target time period, the historical waiting time of the elevator on each floor in the first historical time period equal to and continuous with the duration of the historical target time period, and the historical waiting time of the elevator on each floor in the historical target time period; The hotel server determines the predicted waiting time for the elevator on each floor according to the first reference waiting time for the elevator on each floor and the second reference waiting time for the elevator on each floor; The hotel server inputs the predicted waiting time for the elevator on each floor and the set of room numbers to which the food items need to be delivered into a pre-trained delivery strategy determination model, and determines a food item delivery strategy that minimizes the total time spent by the N delivery robots after all the food items to be delivered are delivered. The delivery strategy determination model is trained by inputting multiple third training samples into a second neural network, each of the third training samples including the historical predicted waiting time for the elevator on each floor, the historical set of room numbers to which the food items need to be delivered, and the corresponding historical food item delivery strategy that minimizes the total time spent after all the food items to be delivered are delivered. The hotel server sends food item delivery instructions to each of the N delivery robots based on the food item delivery strategy with the shortest total delivery time, wherein each food item delivery instruction includes the number of at least one storage space of the food item to be delivered in the food item temporary storage cabinet and the number of at least one room to be delivered; Based on the received food item delivery instructions, the N delivery robots respectively take out the food items associated with the food item delivery instructions from at least one storage space of the food item temporary storage cabinet in turn, and take the elevator in turn to deliver the at least one food item taken out to the door of the guest room corresponding to at least one room number associated with the food item delivery instruction.

2. The method according to claim 1, characterized in that The hotel server determines, based on the voice information of the plurality of guest room telephones, a set of identifications of the guest room telephones of the M guest rooms to which the delivery robot is required to deliver food and beverage items, including: The hotel server converts the voice information of the plurality of guest room telephones into corresponding text information; The hotel server compares each text message with each standard text message in a preset standard text library for requesting food and beverage delivery for similarity; When the similarity between the text message and any standard text message is greater than a set similarity threshold, the hotel server determines that the guest room corresponding to the text message and the guest room phone number thereof require the delivery robot to deliver food and beverage items; The hotel server aggregates the identifiers of multiple guest room telephone numbers that require the delivery robot to deliver food and beverage items, and obtains a set of identifiers of the guest room telephone numbers of M guest rooms that require the delivery robot to deliver food and beverage items.

3. The method according to claim 1, characterized in that The hotel server determines, based on the voice information of the plurality of guest room telephones, a set of identifications of the guest room telephones of the M guest rooms to which the delivery robot is required to deliver food and beverage items, including: The hotel server inputs the voice information of the plurality of guest room telephones into a preset semantic recognition model to respectively identify the semantics corresponding to the voice information of the plurality of guest room telephones; The hotel server determines that the guest room corresponding to the voice information of any guest room phone requires the delivery robot to deliver food and beverage items when the semantics corresponding to the voice information of any guest room phone indicate that the delivery robot is required to deliver food and beverage items; The hotel server aggregates the identifiers of multiple guest room telephone numbers that require the delivery robot to deliver food and beverage items, and obtains a set of identifiers of the guest room telephone numbers of M guest rooms that require the delivery robot to deliver food and beverage items.

4. The method according to claim 1, wherein The hotel server determines the predicted waiting time for the elevator on each floor according to the first reference waiting time for the elevator on each floor and the second reference waiting time for the elevator on each floor, including: The hotel server is based on the formula T 11 =K1T1+K2T2, determine the predicted waiting time for the elevator on each floor, where T 11 is the predicted waiting time for the elevator on any floor, T1 is the first reference waiting time for the elevator on any floor, T2 is the second reference waiting time for the elevator on any floor, K1 is the first weighting coefficient, and K2 is the second weighting coefficient.

5. The method according to claim 1, characterized in that The delivery robot includes a manipulator and a placement tray. The N delivery robots, based on a received food item delivery instruction, sequentially take out food items associated with the food item delivery instruction from at least one storage space of the food item temporary storage cabinet, and sequentially take an elevator to sequentially deliver the taken out at least one food item to a guest room door corresponding to at least one room number associated with the food item delivery instruction, including: For any one of the N delivery robots, receiving a food item delivery instruction; Identifying the number of at least one storage space of the food item temporary storage cabinet and at least one room number to be delivered in the food item delivery instruction; Move to a side of the storage space corresponding to the number of the at least one storage space; Extending a robotic arm to grab a food item to be delivered located in at least one storage space corresponding to the number, and placing at least one food item to be delivered on the placement tray; Generate a navigation route based on the address of the food item storage cabinet and the address associated with at least one room number to be delivered; Based on the navigation route, the customer takes the elevator with the food and beverage items to be delivered and moves to the door of at least one guest room associated with the room number to be delivered.

6. A robot intelligent control device based on speech recognition, characterized in that: The system comprises a hotel server, N delivery robots, multiple guest room telephones, and a temporary storage cabinet for food and beverage items. The temporary storage cabinet for food and beverage items comprises multiple storage spaces, each storage space comprises a unique number, each delivery robot and each guest room telephone are respectively connected to the hotel server for communication, wherein: The hotel server is configured to collect, during a target period of time, voice information transmitted by the plurality of guest room phones to the front desk phone when the guest room phones make calls to the front desk phone; The hotel server is further configured to determine, based on the voice information of the plurality of guest room telephones, a set of identifiers of guest room telephones of M guest rooms to which the delivery robot is required to deliver food and beverage items, where M and N are both positive integers, and M is greater than N; The hotel server is further configured to determine the guest identity information associated with the room phone identifier of each room to which the delivery robot is required to deliver food and beverage items; The hotel server is further configured to obtain order delivery completion information from the online shopping application account associated with the guest identity information, wherein the order delivery completion information includes the room number to which the food and beverage item needs to be delivered and the storage space number of the food and beverage item temporary storage cabinet to be delivered, wherein the room number carries floor information; The hotel server is further configured to input the recorded hotel occupancy volume within the target time period and the number of guest rooms requiring food and beverage delivery by the delivery robot into a pre-trained first elevator waiting time prediction model to determine a first reference waiting time for the elevator on each floor, wherein the elevator waiting time prediction model is trained by inputting a plurality of first training samples into a first neural network, each of the first training samples including the historical hotel occupancy volume within the target time period, the historical number of guest rooms requiring food and beverage delivery by the delivery robot, and the corresponding historical waiting time for the elevator on each floor; The hotel server is further configured to input the actual waiting time for the elevator on each floor in the two previous historical time periods equal to the duration of the target time period, and the actual waiting time for the elevator on each floor in the previous historical time period equal to and continuous with the duration of the target time period, into a pre-trained second elevator waiting time prediction model to determine a second reference waiting time for the elevator on each floor, wherein the second elevator waiting time prediction model is trained by inputting a plurality of second training samples into a long short-term memory model, each of the second training samples including the historical waiting time for the elevator on each floor in the two previous historical time periods equal to the duration of the historical target time period, the historical waiting time for the elevator on each floor in the previous historical time period equal to and continuous with the duration of the historical target time period, and the historical waiting time for the elevator on each floor in the historical target time period; The hotel server is further configured to determine a predicted waiting time for an elevator on each floor based on a first reference waiting time for an elevator on each floor and a second reference waiting time for an elevator on each floor; The hotel server is further configured to input the predicted waiting time for the elevator on each floor and the set of room numbers to which the food items need to be delivered into a pre-trained delivery strategy determination model, and determine a food item delivery strategy that minimizes the total time spent by the N delivery robots after all the food items to be delivered are delivered. The delivery strategy determination model is trained by inputting multiple third training samples into a second neural network, each of the third training samples including the historical predicted waiting time for the elevator on each floor, the historical set of room numbers to which the food items need to be delivered, and the corresponding historical food item delivery strategy that minimizes the total time spent after all the food items to be delivered are delivered. The hotel server is further configured to send food item delivery instructions to each of the N delivery robots based on the food item delivery strategy with the shortest total delivery time, wherein each food item delivery instruction includes the number of at least one storage space of the food item to be delivered in the food item temporary storage cabinet and the number of at least one room to be delivered; The N delivery robots are respectively configured to, based on a received food item delivery instruction, sequentially take out food items associated with the food item delivery instruction from at least one storage space of the food item temporary storage cabinet; The N delivery robots are further configured to take the elevator in turn to deliver the taken-out at least one food and beverage item to the door of the guest room corresponding to at least one room number associated with the food and beverage item delivery instruction.

7. The device according to claim 6, characterized in that The hotel server is specifically used to convert the voice information of multiple guest room telephones into corresponding text information; for each text message, the text information is compared with each standard text message in a preset standard text library for requesting the delivery of food and beverage items; when the similarity between the text information and any standard text information is greater than a set similarity threshold, it is determined that the guest room where the guest room telephone corresponding to the text information is located needs the delivery robot to deliver food and beverage items; the identifiers of multiple guest room telephones that need the delivery robot to deliver food and beverage items are aggregated to obtain a set of identifiers of the guest room telephones of M guest rooms that need the delivery robot to deliver food and beverage items.

8. The device according to claim 6, characterized in that The hotel server is specifically used to input the voice information of multiple guest room telephones into a preset semantic recognition model to respectively identify the semantics corresponding to the voice information of multiple guest room telephones; when the semantics corresponding to the voice information of any guest room telephone indicates that the delivery robot needs to deliver food and beverage items, it is determined that the guest room where the guest room telephone corresponding to the voice information is located needs the delivery robot to deliver food and beverage items; and the identifiers of multiple guest room telephones that need the delivery robot to deliver food and beverage items are aggregated to obtain a set of identifiers of the guest room telephones of M guest rooms that need the delivery robot to deliver food and beverage items.

9. The device according to claim 6, characterized in that The hotel server is specifically used to calculate the 11 =K1T1+K2T2, determine the predicted waiting time for the elevator on each floor, where T 11 is the predicted waiting time for the elevator on any floor, T1 is the first reference waiting time for the elevator on any floor, T2 is the second reference waiting time for the elevator on any floor, K1 is the first weighting coefficient, and K2 is the second weighting coefficient.

10. The device according to claim 6, characterized in that Each of the delivery robots includes a manipulator and a placement tray. Each of the N delivery robots is configured to receive a food item delivery instruction; identify the number of at least one storage space of the food item temporary storage cabinet and the number of at least one room to be delivered in the food item delivery instruction; move sequentially to one side of the storage space corresponding to the number of the at least one storage space; extend the manipulator to grab the food item to be delivered located in the at least one storage space corresponding to the number, and place the at least one food item to be delivered on the placement tray; and generate a navigation route based on the address of the food item temporary storage cabinet and the address associated with the at least one room number to be delivered. Based on the navigation route, the customer takes the elevator with the food and beverage items to be delivered and moves to the door of at least one guest room associated with the room number to be delivered.

Images