Automatic classification and execution method and system for cabin service based on language recognition

Abstract

The present invention relates to the technical field of language recognition, and provides a method and system for automatic classification and execution of cabin services based on language recognition, which monitors the voice messages of all call terminals in the cabin to generate feature information, distributes and connects the call terminals to corresponding message transfer terminals, and forms a voice message queue with the received voice messages; performs first language text recognition on the voice message queue to obtain text content feature information of all voice messages, identifies the service robots in the cabin, and determines the service robots to which the voice message queue is allowed to be sent; performs second language text recognition on the voice message queue to obtain message data feature information, thereby selecting a matching communication channel to transmit the voice message queue to the service robot and adding a source tag to the voice message; generates task instructions based on the voice information received by the service robot, and adjusts the execution status of the task instructions based on the time attribute information of the voice message, so as to realize flexible and scalable voice interactive cabin services.

Description

Technical Field

[0001] The present invention relates to the technical field of language recognition, and in particular to a method and system for automatically classifying and executing cabin services based on language recognition. Background Art

[0002] High-speed rail has become the preferred means of transportation for passengers. The development of high-speed rail is reflected in both hardware and software levels. The riding hardware environment inside the high-speed rail can provide passengers with a spacious and comfortable riding experience. In order to further improve the automated and intelligent services of high-speed rail, cabin services such as catering delivery using robots have emerged. Existing cabin services are based on passengers operating the corresponding service demand buttons on their seats to generate service requests. The train management platform will identify and process service requests from different seats, thereby controlling the robots in the cabin to provide catering delivery services to passengers on the corresponding seats, effectively reducing the labor cost of cabin services. However, existing cabin services can only interact through service demand buttons, reducing the flexibility and scalability of cabin services, and the train management platform usually arranges robots to perform service operations in sequence according to the initiation time of service requests, resulting in robots being able to perform only one service operation during a single trip, which cannot fully utilize the robot's work potential and reduces the efficiency of cabin service execution. Summary of the invention

[0003] In view of the defects of the prior art, the present invention provides a method and system for automatic classification and execution of cabin services based on language recognition, which monitors the voice messages of all call ends in the cabin to generate characteristic information, allocates and connects the call ends to the corresponding message transfer ends, thereby forming a voice message queue with the received voice messages, and realizing centralized management of all voice messages; performs first language text recognition on the voice message queue to obtain text content characteristic information of all voice messages, thereby identifying the service robot in the cabin, determining the service robot to which the voice message queue is allowed to be sent, and designating the service robot to execute the cabin service corresponding to the voice message queue; performs second language text recognition on the voice message queue to obtain message data characteristic information, thereby selecting a matching communication channel to transmit the voice message queue to the service robot and adding a source tag to the voice message, thereby ensuring that the voice message is efficiently sent to the service robot; generates task instructions based on the voice information received by the service robot, and adjusts the execution status of the task instructions based on the time attribute information of the voice message, thereby realizing flexible and scalable voice interactive cabin services and improving the efficiency of cabin service execution.

[0004] The present invention provides a method for automatic classification and execution of cabin services based on language recognition, comprising the following steps:

[0005] Step S1, monitoring all call ends in the cabin to obtain voice message generation feature information of all call ends; based on the voice message generation feature information, all call ends are assigned and connected to corresponding message transfer ends, and the voice messages received by the message transfer ends are formed into a voice message queue;

[0006] Step S2, performing first language text recognition on the voice message queue to obtain text content feature information of all voice messages under the voice message queue; based on the text content feature information, identifying the service robot in the cabin to determine the service robot to which the voice message queue is allowed to be sent;

[0007] Step S3, performing second language text recognition on the voice message queue to obtain message data feature information of all voice messages under the voice message queue; based on the message data feature information, selecting a matching communication channel to transmit the voice message queue from the message transfer terminal to the service robot, and adding source tags to all voice messages under the voice message queue;

[0008] Step S4, based on the voice message received by the service robot, generating a task instruction corresponding to the voice message; based on the time attribute information of all received voice messages, adjusting the execution status of the task instructions corresponding to all received voice messages in the service robot.

[0009] In one embodiment disclosed in the present application, in the step S1, all call terminals in the cabin are monitored to obtain voice message generation feature information of all call terminals; based on the voice message generation feature information, all call terminals are assigned and connected to corresponding message transfer terminals, and the voice messages received by the message transfer terminals are formed into a voice message queue, including:

[0010] Perform visual monitoring on the seats where all call terminals in the cabin are located to obtain the occupancy status information of the seats; based on the occupancy status information, identify all call terminals that are allowed to use the seats; continuously monitor the voice messages of all call terminals that are allowed to use the seats to obtain the voice message generation characteristic information of all the call terminals that are allowed to use the seats; wherein the voice message generation characteristic information includes the characteristic information of the change of the amount of voice message data received by the call terminal over time;

[0011] Based on the voice message generation characteristic information, predict the increasing trend information of the voice message data volume received by the calling end in the future time period; compare the increasing trend information of the voice message data volume of all calling ends with the receivable voice message data volume of all message transfer ends, and determine to assign each calling end to the corresponding message transfer end; based on the chronological order of generation time of all voice messages received by the message transfer end, arrange all voice messages to form a voice message queue.

[0012] In one embodiment disclosed in the present application, in the step S2, the voice message queue is subjected to first language text recognition to obtain text content feature information of all voice messages under the voice message queue; based on the text content feature information, the service robot in the cabin is identified to determine the service robot to which the voice message queue is allowed to be sent, including:

[0013] Based on the message lengths of all the voice messages under the voice message queue, language text recognition with differentiated sampling frequencies is performed on all the voice messages to obtain text keyword content information of all the voice messages under the voice message queue, which is used as the text content feature information;

[0014] The text keyword content information is compared with the service robot operation instruction list to determine the matching degree between all the service robots in the cabin and the voice message queue; based on the matching degree, some service robots in the cabin are identified to determine the service robots to which the voice message queue is allowed to be sent.

[0015] In one embodiment disclosed in the present application, in the step S3, the second language text recognition is performed on the voice message queue to obtain the message data feature information of all voice messages under the voice message queue; based on the message data feature information, a matching communication channel is selected to transmit the voice message queue from the message transfer terminal to the service robot, and a source tag is added to all voice messages under the voice message queue, including:

[0016] Performing language text data code recognition on all voice messages under the voice message queue to obtain message data code quantities of all voice messages under the voice message queue as message data feature information;

[0017] The message data code amount is compared with the available data transmission bandwidth of all idle communication channels in the cabin, and one of the matching communication channels is selected to connect the message transfer end with the service robot, thereby transmitting the voice message queue from the message transfer end to the service robot, and adding the location label of the corresponding source call end in the cabin to all voice messages under the voice message queue as the source label.

[0018] In one embodiment disclosed in the present application, in step S4, based on the voice message received by the service robot, a task instruction corresponding to the voice message is generated; based on the time attribute information of all received voice messages, the execution state of the task instructions corresponding to all received voice messages in the service robot is adjusted, including:

[0019] Parsing and converting the voice message received by the service robot to generate a task instruction corresponding to the voice message; and adding a corresponding execution location tag to the task instruction based on the source tag of the voice message received by the service robot;

[0020] Based on the generation time sequence information of all received voice messages and the execution location location labels of the task instructions corresponding to the received voice messages, the task instructions corresponding to all received voice messages are adjusted to minimize the execution time, thereby determining the execution order of the task instructions corresponding to all received voice messages by the service robot.

[0021] The present invention also provides a cabin service automatic classification and execution system based on language recognition, comprising:

[0022] The call end monitoring and pairing module is used to monitor all call ends in the cabin and obtain the voice message generation feature information of all call ends; based on the voice message generation feature information, all call ends are assigned and connected to the corresponding message transfer end;

[0023] A voice message queue generation module, used for forming a voice message queue from the voice messages received by the message transfer terminal;

[0024] A first language text recognition module, used to perform first language text recognition on the voice message queue to obtain text content feature information of all voice messages under the voice message queue;

[0025] a service robot determination module, configured to identify the service robot in the cabin based on the text content feature information, and determine the service robot to which the voice message queue is allowed to be sent;

[0026] A second language text recognition module, used to perform second language text recognition on the voice message queue to obtain message data feature information of all voice messages under the voice message queue;

[0027] A voice message queue transmission module, used for selecting a matching communication channel to transmit the voice message queue from the message transfer terminal to the service robot based on the message data feature information, and adding a source tag to all voice messages under the voice message queue;

[0028] A task instruction generating module, used for generating a task instruction corresponding to the voice message based on the voice message received by the service robot;

[0029] The task instruction execution adjustment module is used to adjust the execution status of the task instructions corresponding to all received voice messages in the service robot based on the time attribute information of all received voice messages.

[0030] In one embodiment disclosed in the present application, the calling end monitoring and pairing module is used to monitor all calling ends in the cabin to obtain voice message generation feature information of all calling ends; based on the voice message generation feature information, all calling ends are assigned and connected to corresponding message transfer ends, including:

[0031] Perform visual monitoring on the seats where all call terminals in the cabin are located to obtain the occupancy status information of the seats; based on the occupancy status information, identify all call terminals that are allowed to use the seats; continuously monitor the voice messages of all call terminals that are allowed to use the seats to obtain the voice message generation characteristic information of all the call terminals that are allowed to use the seats; wherein the voice message generation characteristic information includes the characteristic information of the change of the amount of voice message data received by the call terminal over time;

[0032] Based on the voice message generation feature information, predict the increase trend information of the voice message data volume received by the calling end in the future time period; compare the increase trend information of the voice message data volume of all calling ends with the receivable voice message data volume of all message transfer ends, and determine to allocate each calling end to the corresponding message transfer end;

[0033] The voice message queue generation module is used to form a voice message queue from the voice messages received by the message transfer terminal, including:

[0034] Based on the generation time sequence of all the voice messages received by the message transfer end, all the voice messages are arranged to form a voice message queue.

[0035] In one embodiment disclosed in the present application, the first language text recognition module is used to perform first language text recognition on the voice message queue to obtain text content feature information of all voice messages under the voice message queue, including:

[0036] Based on the message lengths of all the voice messages under the voice message queue, language text recognition with differentiated sampling frequencies is performed on all the voice messages to obtain text keyword content information of all the voice messages under the voice message queue, which is used as the text content feature information;

[0037] The service robot determination module is used to identify the service robot in the cabin based on the text content feature information and determine the service robot to which the voice message queue is allowed to be sent, including:

[0038] The text keyword content information is compared with the service robot operation instruction list to determine the matching degree between all the service robots in the cabin and the voice message queue; based on the matching degree, some service robots in the cabin are identified to determine the service robots to which the voice message queue is allowed to be sent.

[0039] In one embodiment disclosed in the present application, the second language text recognition module is used to perform second language text recognition on the voice message queue to obtain message data feature information of all voice messages under the voice message queue, including:

[0040] Performing language text data code recognition on all voice messages under the voice message queue to obtain message data code quantities of all voice messages under the voice message queue as message data feature information;

[0041] The voice message queue transmission module is used to select a matching communication channel to transmit the voice message queue from the message transfer terminal to the service robot based on the message data feature information, and add a source tag to all voice messages under the voice message queue, including:

[0042] The message data code amount is compared with the available data transmission bandwidth of all idle communication channels in the cabin, and one of the matching communication channels is selected to connect the message transfer end with the service robot, thereby transmitting the voice message queue from the message transfer end to the service robot, and adding the location label of the corresponding source call end in the cabin to all voice messages under the voice message queue as the source label.

[0043] In one embodiment disclosed in the present application, the task instruction generation module is used to generate a task instruction corresponding to the voice message based on the voice message received by the service robot, including:

[0044] Parsing and converting the voice message received by the service robot to generate a task instruction corresponding to the voice message; and adding a corresponding execution location tag to the task instruction based on the source tag of the voice message received by the service robot;

[0045] The task instruction execution adjustment module is used to adjust the execution status of the task instructions corresponding to all received voice messages in the service robot based on the time attribute information of all received voice messages, including:

[0046] Based on the generation time sequence information of all received voice messages and the execution location location labels of the task instructions corresponding to the received voice messages, the task instructions corresponding to all received voice messages are adjusted to minimize the execution time, thereby determining the execution order of the task instructions corresponding to all received voice messages by the service robot.

[0047] Compared with the prior art, the method and system for automatic classification and execution of cabin services based on language recognition monitors the voice message generation feature information of all call ends in the cabin, distributes and connects the call ends to the corresponding message transfer ends, thereby forming the received voice messages into a voice message queue, and realizing centralized management of all voice messages; performs first language text recognition on the voice message queue to obtain text content feature information of all voice messages, thereby identifying the service robot in the cabin, determining the service robot to which the voice message queue is allowed to be sent, and designating the service robot to execute the cabin service corresponding to the voice message queue; performs second language text recognition on the voice message queue to obtain message data feature information, thereby selecting a matching communication channel to transmit the voice message queue to the service robot and adding a source tag to the voice message, thereby ensuring that the voice message is efficiently sent to the service robot; generates task instructions based on the voice information received by the service robot, and adjusts the execution status of the task instructions based on the time attribute information of the voice message, thereby realizing flexible and scalable voice interactive cabin services and improving the efficiency of cabin service execution.

[0048] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description, claims, and drawings.

[0049] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0051] Figure 1 A flow chart of the method for automatic classification and execution of cabin services based on language recognition provided by the present invention;

[0052] Figure 2 A schematic diagram of the framework of the automatic cabin service classification and execution system based on language recognition provided by the present invention. DETAILED DESCRIPTION

[0053] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0054] See also Figure 1 , is a flow chart of a method for automatically classifying and executing cabin services based on language recognition provided by an embodiment of the present invention. The method for automatically classifying and executing cabin services based on language recognition includes:

[0055] Step S1, monitoring all call ends in the cabin to obtain voice message generation feature information of all call ends; based on the voice message generation feature information, all call ends are assigned and connected to corresponding message transfer ends, and the voice messages received by the message transfer ends are formed into a voice message queue;

[0056] Step S2, performing first language text recognition on the voice message queue to obtain text content feature information of all voice messages under the voice message queue; based on the text content feature information, identifying the service robot in the cabin to determine the service robot to which the voice message queue is allowed to be sent;

[0057] Step S3, performing second language text recognition on the voice message queue to obtain message data feature information of all voice messages under the voice message queue; based on the message data feature information, selecting a matching communication channel to transmit the voice message queue from the message transfer terminal to the service robot, and adding source tags to all voice messages under the voice message queue;

[0058] Step S4, based on the voice message received by the service robot, generating a task instruction corresponding to the voice message; based on the time attribute information of all received voice messages, adjusting the execution status of the task instructions corresponding to all received voice messages in the service robot.

[0059] The method for automatic classification and execution of cabin services based on language recognition monitors the voice message generation feature information of all call ends in the cabin, distributes and connects the call ends to the corresponding message transfer ends, thereby forming the received voice messages into a voice message queue, and realizing centralized management of all voice messages; performs first language text recognition on the voice message queue to obtain text content feature information of all voice messages, thereby identifying the service robot in the cabin, determining the service robot to which the voice message queue is allowed to be sent, and designating the service robot to execute the cabin service corresponding to the voice message queue; performs second language text recognition on the voice message queue to obtain message data feature information, thereby selecting a matching communication channel to transmit the voice message queue to the service robot and adding a source tag to the voice message, thereby ensuring that the voice message is efficiently sent to the service robot; generates task instructions based on the voice information received by the service robot, and adjusts the execution status of the task instructions based on the time attribute information of the voice message, thereby realizing flexible and scalable voice interactive cabin services and improving the efficiency of cabin service execution.

[0060] Preferably, in step S1, all calling terminals in the cabin are monitored to obtain voice message generation feature information of all calling terminals; based on the voice message generation feature information, all calling terminals are assigned and connected to corresponding message transfer terminals, and the voice messages received by the message transfer terminals are formed into a voice message queue, including:

[0061] Perform visual monitoring on the seats where all call terminals in the cabin are located to obtain the occupancy status information of the seats; based on the occupancy status information, identify all call terminals that are allowed to use the seats; continuously monitor the voice messages of all call terminals that are allowed to use the seats to obtain the voice message generation characteristic information of all the call terminals that are allowed to use the seats; wherein the voice message generation characteristic information includes the characteristic information of the change of the amount of voice message data received by the call terminal over time;

[0062] Based on the voice message generation characteristic information, predict the increasing trend information of the voice message data volume received by the call end in the future time period; compare the increasing trend information of the voice message data volume of all call ends with the receivable voice message data volume of all message transfer ends, and determine to assign each call end to the corresponding message transfer end; based on the chronological order of generation time of all voice messages received by the message transfer end, arrange all voice messages to form a voice message queue.

[0063] In the above technical solution, all seats in the train cabin are equipped with call terminals such as voice call devices. Passengers make voice calls to the call terminals, such as issuing agreed voice commands or conducting artificial voice conversations, to trigger the generation of corresponding cabin service voice requests. The train cabin is usually not full, and the cabin service is carried out based on the seats where the passengers are located. If other passengers generate voice calls through the call terminals corresponding to the seats where no passengers are arranged, it will cause the cabin service to be executed incorrectly, seriously wasting the cabin service execution power. In order to avoid the abuse of seats where no passengers are arranged, the seats where all call terminals are located in the cabin are first visually identified and monitored to obtain the appearance feature information of the people on the seats, and the appearance feature information of the people is compared with the appearance feature information of the people who purchase tickets for the cabin seats. If the two are consistent, it is determined that the call terminal of the corresponding seat is in a state of being allowed to be used; if the two are inconsistent, it is determined that the call terminal of the corresponding seat is not in a state of being allowed to be used, and the call terminal that is not in a state of being allowed to be used is switched to a disabled state to avoid the abuse of the call terminal and affect the normal execution of the cabin service. Then, the calling end that is allowed to be used is continuously monitored to obtain characteristic information of the change in the amount of voice message data received by the calling end that is allowed to be used during the monitoring period over time (i.e., characteristic information of the change in the amount of received voice message data over time), thereby dynamically identifying the frequency of voice calls of the calling end that is allowed to be used. The method also uses the characteristic information of the amount of voice message data received by the calling end in the monitoring period as a benchmark, predicts the increasing trend of the amount of voice message data received by the calling end in the future time period, and quantitatively predicts the amount of voice message data received by the calling end in the future time period; then compares the increasing trend information of the amount of voice message data of all calling ends with the amount of voice message data that can be received by all message transfer ends, that is, determines the size relationship between the amount of voice message data that can be received by each message transfer end and the increasing trend of the amount of voice message data of each calling end. If the amount of voice message data that can be received by a certain message transfer end is less than the increase value of the amount of voice message data of a certain calling end in the future preset time end, it is determined that the message transfer end cannot be docked with the calling end; otherwise, it is determined that the message transfer end can be docked with the calling end, so as to allocate a suitable message transfer end for docking to each calling end, ensuring that the voice messages generated by each calling end can be accurately sent and processed. Based on the generation time sequence of all voice messages received by the message transfer end, all voice messages are arranged to form a voice message queue, ensuring that all voice messages are processed in time.

[0064] Preferably, in step S2, the voice message queue is subjected to first language text recognition to obtain text content feature information of all voice messages under the voice message queue; based on the text content feature information, the service robot in the cabin is identified to determine the service robot to which the voice message queue is allowed to be sent, including:

[0065] Based on the message lengths of all the voice messages under the voice message queue, language text recognition with differentiated sampling frequencies is performed on all the voice messages to obtain text keyword content information of all the voice messages under the voice message queue, which is used as the text content feature information;

[0066] The keyword content information of the text is compared with the service robot operation instruction list to determine the matching degree between all the service robots in the cabin and the voice message queue; based on the matching degree, some service robots in the cabin are identified to determine the service robots to which the voice message queue is allowed to be sent.

[0067] In the above technical solution, the voice message queue includes multiple voice messages, each voice message is sent by the calling end to the corresponding passenger, and the length of the voice message matches the passenger's cabin service demand. The more complex the passenger's cabin service demand is, the longer the corresponding voice message is. In order to perform comprehensive voice-text recognition on each voice message, the message length of all voice messages is used as a benchmark, and language text recognition with differentiated sampling frequency is performed on all voice messages. The longer the message length of the voice message is, the greater the sampling frequency of the voice message is, and the text keyword content information of all voice messages under the voice message queue (that is, all text keywords contained in the voice message) is obtained, thereby characterizing the text content of all voice messages and comprehensively determining the passenger's voice call cabin service demand. In addition, the types of cabin service operations that each service robot in the cabin is allowed to perform are fixed, and not all service robots can respond to the cabin service needs of all passengers. To ensure that each passenger obtains matching cabin services, the text keyword content information is compared with the service robot operation instruction list to determine the matching degree between all service robots in the cabin and the voice message queue, wherein the service robot operation instruction list includes control trigger instructions corresponding to all cabin services that can be performed by the service robot. The more the text keyword content information matches the control trigger instructions contained in the service robot operation instruction list, the higher the matching degree between the two. Based on the matching degree, at least one service robot that can execute the maximum number of cabin service needs corresponding to all voice messages under the voice message queue is determined, and the voice message queue is sent to the at least one service robot, so that the service robot can respond to the voice message and perform the corresponding cabin service.

[0068] Preferably, in step S3, the second language text recognition is performed on the voice message queue to obtain message data feature information of all voice messages under the voice message queue; based on the message data feature information, a matching communication channel is selected to transmit the voice message queue from the message transfer terminal to the service robot, and source tags are added to all voice messages under the voice message queue, including:

[0069] Perform language text data code recognition on all voice messages under the voice message queue to obtain message data code quantities of all voice messages under the voice message queue as message data feature information;

[0070] The message data code amount is compared with the available data transmission bandwidth of all idle communication channels in the cabin, and one of the matching communication channels is selected to connect the message transfer end with the service robot, thereby transmitting the voice message queue from the message transfer end to the service robot, and adding the location label of the corresponding source call end in the cabin to all voice messages under the voice message queue as the source label.

[0071] In the above technical solution, several service robots are distributed in the cabin, and the service robots are connected to the message transfer terminal through a wireless communication channel. If the data transmission bandwidth of the wireless communication channel is too small, it is easy to cause congestion in the voice message queue during the transmission from the message transfer terminal to the service robot, and it is impossible to ensure that the voice message queue is sent to the service robot in time, which reduces the service robot's response speed to voice messages and the efficiency of cabin service execution. To this end, language text data codes are identified for all voice messages under the voice message queue to obtain the message data code quantities of all voice messages under the voice message queue, and the message data code quantities are compared with the available data transmission bandwidths of all idle communication channels in the cabin, and one of the communication channels compatible with the transmission of the message data code quantity is selected to connect the message transfer end with the service robot (for example, the communication channel with the largest available data transmission bandwidth is selected to connect the message transfer end with the service robot), thereby transmitting the voice message queue from the message transfer end to the service robot, and adding location tags of the corresponding source call ends in the cabin to all voice messages under the voice message queue, so as to locate the source of the cabin service demand corresponding to each voice message, and ensure that the service robot can move to the corresponding position in the cabin to perform cabin service.

[0072] Preferably, in step S4, based on the voice message received by the service robot, a task instruction corresponding to the voice message is generated; based on the time attribute information of all received voice messages, the execution state of the task instructions corresponding to all received voice messages in the service robot is adjusted, including:

[0073] Parse and convert the voice message received by the service robot to generate a task instruction corresponding to the voice message; and add a corresponding execution location tag to the task instruction based on the source tag of the voice message received by the service robot;

[0074] Based on the generation time sequence information of all received voice messages and the execution location location labels of the task instructions corresponding to the received voice messages, the task instructions corresponding to all received voice messages are adjusted to minimize the execution time, thereby determining the execution order of the task instructions corresponding to all received voice messages by the service robot.

[0075] In the above technical solution, the voice message received by the service robot is parsed and converted to generate a task instruction corresponding to the voice message, and based on the source tag of the voice message received by the service robot, a corresponding execution location tag is added to the task instruction, so as to accurately control the service robot to move to the corresponding seat in the cabin to perform cabin service. In addition, based on the generation time sequence information of all received voice messages and the execution location tag of the task instruction corresponding to the received voice message, the service robot moves in sequence in the cabin to all cabin service execution locations to complete all cabin services in the shortest time corresponding motion path, so as to determine the execution order of the task instructions corresponding to all received voice messages by the service robot, and improve the efficiency of cabin service execution.

[0076] See also Figure 2 , is a schematic diagram of the framework of a cabin service automation classification and execution system based on language recognition provided by an embodiment of the present invention. The cabin service automation classification and execution system based on language recognition includes:

[0077] The call end monitoring and pairing module is used to monitor all call ends in the cabin and obtain the voice message generation feature information of all call ends; based on the voice message generation feature information, all call ends are assigned and connected to the corresponding message transfer end;

[0078] A voice message queue generation module is used to form a voice message queue with the voice messages received by the message transfer end;

[0079] A first language text recognition module, used to perform first language text recognition on the voice message queue to obtain text content feature information of all voice messages under the voice message queue;

[0080] A service robot determination module, used to identify the service robot in the cabin based on the text content feature information, and determine the service robot to which the voice message queue is allowed to be sent;

[0081] A second language text recognition module is used to perform second language text recognition on the voice message queue to obtain message data feature information of all voice messages under the voice message queue;

[0082] A voice message queue transmission module is used to select a matching communication channel to transmit the voice message queue from the message transfer terminal to the service robot based on the message data feature information, and add a source tag to all voice messages under the voice message queue;

[0083] A task instruction generating module, used for generating a task instruction corresponding to the voice message based on the voice message received by the service robot;

[0084] The task instruction execution adjustment module is used to adjust the execution status of the task instructions corresponding to all received voice messages in the service robot based on the time attribute information of all received voice messages.

[0085] The cabin service automation classification and execution system based on language recognition monitors the voice message generation feature information of all call ends in the cabin, distributes and connects the call ends to the corresponding message transfer ends, thereby forming the received voice messages into a voice message queue, and realizing centralized management of all voice messages; performs first language text recognition on the voice message queue to obtain text content feature information of all voice messages, thereby identifying the service robot in the cabin, determining the service robot to which the voice message queue is allowed to be sent, and designating the service robot to execute the cabin service corresponding to the voice message queue; performs second language text recognition on the voice message queue to obtain message data feature information, thereby selecting a matching communication channel to transmit the voice message queue to the service robot and adding a source tag to the voice message, thereby ensuring that the voice message is efficiently sent to the service robot; generates task instructions based on the voice information received by the service robot, and adjusts the execution status of the task instructions based on the time attribute information of the voice message, thereby realizing flexible and scalable voice interactive cabin services and improving the efficiency of cabin service execution.

[0086] Preferably, the call end monitoring and pairing module is used to monitor all call ends in the cabin to obtain voice message generation feature information of all call ends; based on the voice message generation feature information, all call ends are assigned and connected to corresponding message transfer ends, including:

[0087] Perform visual monitoring on the seats where all call terminals in the cabin are located to obtain the occupancy status information of the seats; based on the occupancy status information, identify all call terminals that are allowed to use the seats; continuously monitor the voice messages of all call terminals that are allowed to use the seats to obtain the voice message generation characteristic information of all the call terminals that are allowed to use the seats; wherein the voice message generation characteristic information includes the characteristic information of the change of the amount of voice message data received by the call terminal over time;

[0088] Based on the voice message generation feature information, predict the voice message data volume increase trend information received by the calling end in the future time period; compare the voice message data volume increase trend information of all calling ends with the receivable voice message data volume of all message transfer ends, and determine to allocate each calling end to the corresponding message transfer end;

[0089] The voice message queue generation module is used to form a voice message queue with the voice messages received by the message transfer end, including:

[0090] Based on the generation time sequence of all the voice messages received by the message transfer end, all the voice messages are arranged to form a voice message queue.

[0091] In the above technical solution, all seats in the train cabin are equipped with call terminals such as voice call devices. Passengers make voice calls to the call terminals, such as issuing agreed voice commands or conducting artificial voice conversations, to trigger the generation of corresponding cabin service voice requests. The train cabin is usually not full, and the cabin service is carried out based on the seats where the passengers are located. If other passengers generate voice calls through the call terminals corresponding to the seats where no passengers are arranged, it will cause the cabin service to be executed incorrectly, seriously wasting the cabin service execution power. In order to avoid the abuse of seats where no passengers are arranged, the seats where all call terminals are located in the cabin are first visually identified and monitored to obtain the appearance feature information of the people on the seats, and the appearance feature information of the people is compared with the appearance feature information of the people who purchase tickets for the cabin seats. If the two are consistent, it is determined that the call terminal of the corresponding seat is in a state of being allowed to be used; if the two are inconsistent, it is determined that the call terminal of the corresponding seat is not in a state of being allowed to be used, and the call terminal that is not in a state of being allowed to be used is switched to a disabled state to avoid the abuse of the call terminal and affect the normal execution of the cabin service. Then, the calling end that is allowed to be used is continuously monitored to obtain characteristic information of the change in the amount of voice message data received by the calling end that is allowed to be used during the monitoring period over time (i.e., characteristic information of the change in the amount of received voice message data over time), thereby dynamically identifying the frequency of voice calls of the calling end that is allowed to be used. The method also uses the characteristic information of the amount of voice message data received by the calling end in the monitoring period as a benchmark, predicts the increasing trend of the amount of voice message data received by the calling end in the future time period, and quantitatively predicts the amount of voice message data received by the calling end in the future time period; then compares the increasing trend information of the amount of voice message data of all calling ends with the amount of voice message data that can be received by all message transfer ends, that is, determines the size relationship between the amount of voice message data that can be received by each message transfer end and the increasing trend of the amount of voice message data of each calling end. If the amount of voice message data that can be received by a certain message transfer end is less than the increase value of the amount of voice message data of a certain calling end in the future preset time end, it is determined that the message transfer end cannot be docked with the calling end; otherwise, it is determined that the message transfer end can be docked with the calling end, so as to allocate a suitable message transfer end for docking to each calling end, ensuring that the voice messages generated by each calling end can be accurately sent and processed. Based on the generation time sequence of all voice messages received by the message transfer end, all voice messages are arranged to form a voice message queue, ensuring that all voice messages are processed in time.

[0092] Preferably, the first language text recognition module is used to perform first language text recognition on the voice message queue to obtain text content feature information of all voice messages under the voice message queue, including:

[0093] Based on the message lengths of all the voice messages under the voice message queue, language text recognition with differentiated sampling frequencies is performed on all the voice messages to obtain text keyword content information of all the voice messages under the voice message queue, which is used as the text content feature information;

[0094] The service robot determination module is used to identify the service robot in the cabin based on the text content feature information, and determine the service robot to which the voice message queue is allowed to be sent, including:

[0095] The keyword content information of the text is compared with the service robot operation instruction list to determine the matching degree between all the service robots in the cabin and the voice message queue; based on the matching degree, some service robots in the cabin are identified to determine the service robots to which the voice message queue is allowed to be sent.

[0096] In the above technical solution, the voice message queue includes multiple voice messages, each voice message is sent by the calling end to the corresponding passenger, and the length of the voice message matches the passenger's cabin service demand. The more complex the passenger's cabin service demand is, the longer the corresponding voice message is. In order to perform comprehensive voice-text recognition on each voice message, the message length of all voice messages is used as a benchmark, and language text recognition with differentiated sampling frequency is performed on all voice messages. The longer the message length of the voice message is, the greater the sampling frequency of the voice message is, and the text keyword content information of all voice messages under the voice message queue (that is, all text keywords contained in the voice message) is obtained, thereby characterizing the text content of all voice messages and comprehensively determining the passenger's voice call cabin service demand. In addition, the types of cabin service operations that each service robot in the cabin is allowed to perform are fixed, and not all service robots can respond to the cabin service needs of all passengers. To ensure that each passenger obtains matching cabin services, the text keyword content information is compared with the service robot operation instruction list to determine the matching degree between all service robots in the cabin and the voice message queue, wherein the service robot operation instruction list includes control trigger instructions corresponding to all cabin services that can be performed by the service robot. The more the text keyword content information matches the control trigger instructions contained in the service robot operation instruction list, the higher the matching degree between the two. Based on the matching degree, at least one service robot that can execute the maximum number of cabin service needs corresponding to all voice messages under the voice message queue is determined, and the voice message queue is sent to the at least one service robot, so that the service robot can respond to the voice message and perform the corresponding cabin service.

[0097] Preferably, the second language text recognition module is used to perform second language text recognition on the voice message queue to obtain message data feature information of all voice messages under the voice message queue, including:

[0098] Perform language text data code recognition on all voice messages under the voice message queue to obtain message data code quantities of all voice messages under the voice message queue as message data feature information;

[0099] The voice message queue transmission module is used to select a matching communication channel to transmit the voice message queue from the message transfer end to the service robot based on the message data feature information, and add a source tag to all voice messages under the voice message queue, including:

[0100] The message data code amount is compared with the available data transmission bandwidth of all idle communication channels in the cabin, and one of the matching communication channels is selected to connect the message transfer end with the service robot, thereby transmitting the voice message queue from the message transfer end to the service robot, and adding the location label of the corresponding source call end in the cabin to all voice messages under the voice message queue as the source label.

[0101] In the above technical solution, several service robots are distributed in the cabin, and the service robots are connected to the message transfer terminal through a wireless communication channel. If the data transmission bandwidth of the wireless communication channel is too small, it is easy to cause congestion in the voice message queue during the transmission from the message transfer terminal to the service robot, and it is impossible to ensure that the voice message queue is sent to the service robot in time, which reduces the service robot's response speed to voice messages and the efficiency of cabin service execution. To this end, language text data codes are identified for all voice messages under the voice message queue to obtain the message data code quantities of all voice messages under the voice message queue, and the message data code quantities are compared with the available data transmission bandwidths of all idle communication channels in the cabin, and one of the communication channels compatible with the transmission of the message data code quantity is selected to connect the message transfer end with the service robot (for example, the communication channel with the largest available data transmission bandwidth is selected to connect the message transfer end with the service robot), thereby transmitting the voice message queue from the message transfer end to the service robot, and adding location tags of the corresponding source call ends in the cabin to all voice messages under the voice message queue, so as to locate the source of the cabin service demand corresponding to each voice message, and ensure that the service robot can move to the corresponding position in the cabin to perform cabin service.

[0102] Preferably, the task instruction generating module is used to generate a task instruction corresponding to the voice message based on the voice message received by the service robot, including:

[0103] Parse and convert the voice message received by the service robot to generate a task instruction corresponding to the voice message; and add a corresponding execution location tag to the task instruction based on the source tag of the voice message received by the service robot;

[0104] The task instruction execution adjustment module is used to adjust the execution status of the task instructions corresponding to all received voice messages in the service robot based on the time attribute information of all received voice messages, including:

[0105] Based on the generation time sequence information of all received voice messages and the execution location location labels of the task instructions corresponding to the received voice messages, the task instructions corresponding to all received voice messages are adjusted to minimize the execution time, thereby determining the execution order of the task instructions corresponding to all received voice messages by the service robot.

[0106] In the above technical solution, the voice message received by the service robot is parsed and converted to generate a task instruction corresponding to the voice message, and based on the source tag of the voice message received by the service robot, a corresponding execution location tag is added to the task instruction, so as to accurately control the service robot to move to the corresponding seat in the cabin to perform cabin service. In addition, based on the generation time sequence information of all received voice messages and the execution location tag of the task instruction corresponding to the received voice message, the service robot moves in sequence in the cabin to all cabin service execution locations to complete all cabin services in the shortest time corresponding motion path, so as to determine the execution order of the task instructions corresponding to all received voice messages by the service robot, and improve the efficiency of cabin service execution.

[0107] It can be seen from the contents of the above embodiments that the method and system for automatic classification and execution of cabin services based on language recognition monitor the voice message generation feature information of all call ends in the cabin, and assign the call end to the corresponding message transfer end, so as to form a voice message queue with the received voice messages, and realize centralized management of all voice messages; perform first language text recognition on the voice message queue to obtain text content feature information of all voice messages, thereby identifying the service robot in the cabin, determining the service robot to which the voice message queue is allowed to be sent, and specifying the service robot to execute the cabin service corresponding to the voice message queue; perform second language text recognition on the voice message queue to obtain message data feature information, thereby selecting a matching communication channel to transmit the voice message queue to the service robot and add a source tag to the voice message to ensure that the voice message is efficiently sent to the service robot; generate task instructions based on the voice information received by the service robot, and adjust the execution status of the task instructions based on the time attribute information of the voice message, so as to realize flexible and scalable voice interactive cabin services and improve the efficiency of cabin service execution.

[0108] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A method for automatic classification and execution of cabin services based on language recognition, characterized in that: It includes the following steps: Step S1, monitor all call terminals in the cabin to obtain voice message generation characteristic information of all call terminals, including visually monitoring the seats where all call terminals are located in the cabin to obtain the occupancy status information of the seats; based on the occupancy status information, identify all call terminals that are allowed to use; continuously monitor voice messages of all call terminals that are allowed to use to obtain the voice message generation characteristic information of all call terminals that are allowed to use; wherein the voice message generation characteristic information includes characteristic information of the amount of voice message data received by the call terminal over time; based on the voice message generation characteristic information, assign and connect all call terminals to corresponding message transfer terminals, and form a voice message queue with the voice messages received by the message transfer terminal; Step S2, performing first language text recognition on the voice message queue to obtain text content feature information of all voice messages under the voice message queue; based on the text content feature information, identifying the service robot in the cabin to determine the service robot to which the voice message queue is allowed to be sent; Step S3, performing second language text recognition on the voice message queue to obtain message data feature information of all voice messages under the voice message queue, including performing language text data code recognition on all voice messages under the voice message queue to obtain message data code amounts of all voice messages under the voice message queue as message data feature information; based on the message data feature information, selecting a matching communication channel to transmit the voice message queue from the message transfer terminal to the service robot, and adding source tags to all voice messages under the voice message queue; Step S4, based on the voice message received by the service robot, generating a task instruction corresponding to the voice message; based on the time attribute information of all received voice messages, adjusting the execution status of the task instructions corresponding to all received voice messages in the service robot.

2. The method for automatic cabin service classification and execution based on language recognition according to claim 1, characterized in that: In the step S1, based on the characteristic information of the voice message generation, all call ends are assigned to corresponding message transfer ends, and the voice messages received by the message transfer ends are formed into a voice message queue, including: Based on the voice message generation characteristic information, predict the increasing trend information of the voice message data volume received by the calling end in the future time period; compare the increasing trend information of the voice message data volume of all calling ends with the receivable voice message data volume of all message transfer ends, and determine to assign each calling end to the corresponding message transfer end; based on the chronological order of generation time of all voice messages received by the message transfer end, arrange all voice messages to form a voice message queue.

3. The method for automatic cabin service classification and execution based on language recognition according to claim 1, characterized in that: In the step S2, the first language text recognition is performed on the voice message queue to obtain text content feature information of all voice messages under the voice message queue; Based on the text content feature information, identifying the service robot in the cabin and determining the service robot to which the voice message queue is allowed to be sent includes: Based on the message lengths of all the voice messages under the voice message queue, language text recognition with differentiated sampling frequencies is performed on all the voice messages to obtain text keyword content information of all the voice messages under the voice message queue, which is used as the text content feature information; The text keyword content information is compared with the service robot operation instruction list to determine the matching degree between all the service robots in the cabin and the voice message queue; based on the matching degree, some service robots in the cabin are identified to determine the service robots to which the voice message queue is allowed to be sent.

4. The method for automatic classification and execution of cabin services based on language recognition according to claim 1, characterized in that: In the step S3, based on the message data feature information, a matching communication channel is selected to transmit the voice message queue from the message transfer terminal to the service robot, and a source tag is added to all voice messages under the voice message queue, including: The message data code amount is compared with the available data transmission bandwidth of all idle communication channels in the cabin, and one of the matching communication channels is selected to connect the message transfer end with the service robot, thereby transmitting the voice message queue from the message transfer end to the service robot, and adding the location label of the corresponding source call end in the cabin to all voice messages under the voice message queue as the source label.

5. The method for automatic cabin service classification and execution based on language recognition according to claim 1, characterized in that: In the step S4, based on the voice message received by the service robot, a task instruction corresponding to the voice message is generated; Based on the time attribute information of all received voice messages, adjusting the execution status of the task instructions corresponding to all received voice messages in the service robot includes: Parsing and converting the voice message received by the service robot to generate a task instruction corresponding to the voice message; and adding a corresponding execution location tag to the task instruction based on the source tag of the voice message received by the service robot; Based on the generation time sequence information of all received voice messages and the execution location location labels of the task instructions corresponding to the received voice messages, the task instructions corresponding to all received voice messages are adjusted to minimize the execution time, thereby determining the execution order of the task instructions corresponding to all received voice messages by the service robot.

6. The cabin service automatic classification and execution system based on language recognition is characterized by: include: The call end monitoring and pairing module is used to monitor all call ends in the cabin and obtain the voice message generation characteristic information of all call ends, including visually monitoring the seats where all call ends in the cabin are located to obtain the occupancy status information of the seats; based on the occupancy status information, identifying all call ends in the allowed use state; continuously monitoring the voice messages of all call ends in the allowed use state to obtain the voice message generation characteristic information of all call ends in the allowed use state; wherein, the voice message generation characteristic information includes the characteristic information of the amount of voice message data received by the call end over time; based on the voice message generation characteristic information, all call ends are assigned and docked to the corresponding message transfer ends; A voice message queue generation module, used for forming a voice message queue from the voice messages received by the message transfer terminal; A first language text recognition module, used to perform first language text recognition on the voice message queue to obtain text content feature information of all voice messages under the voice message queue; a service robot determination module, configured to identify the service robot in the cabin based on the text content feature information, and determine the service robot to which the voice message queue is allowed to be sent; A second language text recognition module is used to perform second language text recognition on the voice message queue to obtain message data feature information of all voice messages under the voice message queue, including performing language text data code recognition on all voice messages under the voice message queue to obtain message data code amounts of all voice messages under the voice message queue as message data feature information; A voice message queue transmission module, used for selecting a matching communication channel to transmit the voice message queue from the message transfer terminal to the service robot based on the message data feature information, and adding a source tag to all voice messages under the voice message queue; A task instruction generating module, used for generating a task instruction corresponding to the voice message based on the voice message received by the service robot; The task instruction execution adjustment module is used to adjust the execution status of the task instructions corresponding to all received voice messages in the service robot based on the time attribute information of all received voice messages.

7. The cabin service automatic classification and execution system based on language recognition according to claim 6, characterized in that: The calling end monitoring and pairing module is used to generate characteristic information based on the voice message and assign all calling ends to corresponding message transfer ends, including: Based on the voice message generation feature information, predict the increase trend information of the voice message data volume received by the calling end in the future time period; compare the increase trend information of the voice message data volume of all calling ends with the receivable voice message data volume of all message transfer ends, and determine to allocate each calling end to the corresponding message transfer end; The voice message queue generation module is used to form a voice message queue from the voice messages received by the message transfer terminal, including: Based on the generation time sequence of all the voice messages received by the message transfer end, all the voice messages are arranged to form a voice message queue.

8. The cabin service automatic classification and execution system based on language recognition according to claim 6, characterized in that: The first language text recognition module is used to perform first language text recognition on the voice message queue to obtain text content feature information of all voice messages under the voice message queue, including: Based on the message lengths of all the voice messages under the voice message queue, language text recognition with differentiated sampling frequencies is performed on all the voice messages to obtain text keyword content information of all the voice messages under the voice message queue, which is used as the text content feature information; The service robot determination module is used to identify the service robot in the cabin based on the text content feature information and determine the service robot to which the voice message queue is allowed to be sent, including: The text keyword content information is compared with the service robot operation instruction list to determine the matching degree between all the service robots in the cabin and the voice message queue; based on the matching degree, some service robots in the cabin are identified to determine the service robots to which the voice message queue is allowed to be sent.

9. The cabin service automatic classification and execution system based on language recognition according to claim 6, characterized in that: The voice message queue transmission module is used to select a matching communication channel to transmit the voice message queue from the message transfer terminal to the service robot based on the message data feature information, and add a source tag to all voice messages under the voice message queue, including: The message data code amount is compared with the available data transmission bandwidth of all idle communication channels in the cabin, and one of the matching communication channels is selected to connect the message transfer end with the service robot, thereby transmitting the voice message queue from the message transfer end to the service robot, and adding the location label of the corresponding source call end in the cabin to all voice messages under the voice message queue as the source label.

10. The cabin service automatic classification and execution system based on language recognition according to claim 6, characterized in that: The task instruction generating module is used to generate a task instruction corresponding to the voice message based on the voice message received by the service robot, including: Parsing and converting the voice message received by the service robot to generate a task instruction corresponding to the voice message; and adding a corresponding execution location tag to the task instruction based on the source tag of the voice message received by the service robot; The task instruction execution adjustment module is used to adjust the execution status of the task instructions corresponding to all received voice messages in the service robot based on the time attribute information of all received voice messages, including: Based on the generation time sequence information of all received voice messages and the execution location location labels of the task instructions corresponding to the received voice messages, the task instructions corresponding to all received voice messages are adjusted to minimize the execution time, thereby determining the execution order of the task instructions corresponding to all received voice messages by the service robot.

Images