JP2013241226A - Elevator call registration device and method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To obtain an elevator call registration device and a method capable of eliminating an unnecessary waiting time while securing a time required for changing call registration by performing the call registration for a destination floor of a passenger through voice recognition.SOLUTION: A matching rate between a voice signal obtained by converting voice generated by a passenger and a reference voice signal corresponding to each destination floor stored in advance is calculated. A specified time which is measured until a call is registered after voice is specified is determined according to a calculated matching rate. When the determined specified time passes, call registration for a destination floor is performed.

Description

  The present invention relates to an elevator call registration apparatus and an elevator call registration method for performing call registration on a destination floor by voice recognition.

  2. Description of the Related Art Conventionally, as an elevator call registration device, there has been known a voice call registration device that enables call registration by recognizing a voice uttered by a passenger (see, for example, Patent Document 1).

  In addition, there is a method for notifying whether or not the passenger has made call registration correctly by displaying the result of voice recognition on a display device (see, for example, Patent Document 2). Furthermore, there is a method of notifying the degree of possibility that the passenger can accurately perform the call registration by performing level discrimination on the input of the audio signal corresponding to the voice emitted by the passenger and displaying the result on the display device. Yes (see, for example, Patent Document 3).

JP-A-1-247378 JP-A-6-255911 International Publication No. 11/088663 Pamphlet

However, the prior art has the following problems.
In the prior art as disclosed in Patent Document 1, a fixed time is provided between voice recognition and call registration so that passengers can change call registration. However, since this fixed time is fixed, there is a case where the call registration cannot be changed in time.

  Therefore, if the passenger's destination floor is mistakenly recognized by voice, the passenger cannot perform the desired call registration. On the other hand, even if the destination floor is correctly recognized by voice, the passenger is recognized by voice. It was necessary to wait for a certain period of time until call registration was confirmed, and an unnecessary waiting time occurred.

  Even in the prior art as shown in Patent Documents 2 and 3, the passenger was able to know the status of call registration, but the fixed time provided between the time of voice recognition and the time of call registration is A similar problem occurred because it was always constant.

  The present invention has been made in order to solve the above-described problems. An elevator call registration device and an elevator that can eliminate unnecessary waiting time while securing the time required for changing call registration. The purpose is to obtain a call registration method.

  The elevator call registration device according to the present invention is connected to an elevator control panel that controls the operation of the elevator according to the registered destination floor, and calls the destination floor with respect to the elevator control panel based on the voice emitted by the passengers. An elevator call registration device that performs registration, in which a reference audio signal corresponding to each destination floor is stored in advance, recognizes the destination floor from an audio signal converted from a voice emitted by a passenger, and the audio signal and the reference audio The degree of coincidence of the audio signal is calculated by comparing with the signal, and according to the calculated degree of coincidence, the voice recognition unit that determines that the recognized destination floor can be registered as a call, and the voice recognition unit determines that the call can be registered. A fixed time table that correlates the fixed time, which is the time from when the destination floor is recognized until the destination floor is called and registered, and the degree of coincidence is stored beforehand. The confirmation time corresponding to the degree of coincidence calculated by the speech recognition means is extracted from the confirmation time table, and when the confirmation time extracted by the confirmation time determination means has elapsed, the speech recognition means Call destination processing means for outputting the destination floor determined to be call-registerable to the elevator control panel as the destination floor to be call-registered and performing call registration of the destination floor is provided.

  Also, the elevator call registration method in the present invention is connected to an elevator control panel that controls the operation of the elevator according to the destination floor registered for call registration, and is based on the voice emitted by the passenger with respect to the elevator control panel. The elevator call registration method executed by the elevator call registration device for performing the call registration, a first storage step for storing in advance a reference voice signal corresponding to each destination floor in the storage unit, and a passenger The destination floor is recognized from the voice signal converted from the voice emitted from the voice signal, and the voice signal coincidence is calculated by comparing the voice signal with the reference voice signal stored in the storage unit. A speech recognition step for determining that the recognized destination floor can be registered by calling, and a destination floor after recognizing the destination floor determined to be call-registerable by the speech recognition step. A second storage step in which a fixed time table that associates a fixed time that is a time until call registration and a matching degree are stored in the storage unit in advance, and a fixed corresponding to the matching degree calculated in the voice recognition step A fixed time determination step for extracting the time from the fixed time table stored in the storage unit, and a destination floor that is determined to be call-registerable by the voice recognition step when the fixed time extracted in the fixed time determination step has elapsed. Is output to the elevator control panel as a destination floor to be registered for call registration, and a call registration processing step for performing call registration for the destination floor is provided.

  According to the elevator call registration device and the elevator call registration method of the present invention, call registration is determined by determining a fixed time, which is a time from voice recognition to call registration, according to the degree of coincidence of voice signals. It is possible to obtain an elevator call registration device and an elevator call registration method that can eliminate unnecessary waiting time while securing the time required for the change.

It is a block diagram of the elevator call registration apparatus in Embodiment 1 of this invention. It is a flowchart which shows the speech recognition process in Embodiment 1 of this invention. It is a flowchart which shows the speech recognition process in Embodiment 2 of this invention. It is a flowchart which shows the speech recognition process in Embodiment 3 of this invention.

  Hereinafter, preferred embodiments of an elevator call registration device and an elevator call registration method according to the present invention will be described with reference to the drawings.

Embodiment 1 FIG.
FIG. 1 is a configuration diagram of an elevator call registration device 10 according to Embodiment 1 of the present invention. A voice input device 20 and an elevator control panel 30 are connected to the call registration device 10 in FIG.

  The voice input device 20 converts a voice uttered by a passenger into a voice signal, and inputs the voice signal to the call registration device 10. The elevator control panel 30 inputs the operation state of the elevator to the call registration device 10 and further controls the operation of the elevator according to the destination floor registered for call.

  For the voice input device 20, for example, a microphone or the like is used. The voice input device 20 may be installed in an elevator car or may be installed in an elevator platform. In the first to third embodiments, the voice input device 20 will be described as being installed in an elevator car.

  Here, the elevator call registration device according to the first embodiment represents a reference for determining whether the destination floor is correctly voice-recognized or wrongly voice-recognized by the degree of coincidence of the sound signals, as compared with the prior art. It has a technical feature of determining an optimum determination time according to the degree of coincidence of audio signals.

  Next, the operation of the call registration device 10 in FIG. 1 will be described. The call registration device 10 includes voice recognition means 11, fixed time determination means 12, and call registration processing means 13.

  The voice recognition unit 11 recognizes the passenger's destination floor from the voice signal input from the voice input device 20. Furthermore, the speech recognition means 11 compares the input speech signal with a reference speech signal corresponding to each destination floor stored in advance in a storage unit (not shown), and determines the degree of coincidence of these speech signals. calculate.

  If the voice recognition unit 11 determines that the recognized destination floor can be registered for call according to the degree of coincidence of the audio signals, the voice recognition unit 11 inputs the recognized destination floor to the call registration processing unit 13. Further, the voice recognition unit 11 inputs the calculated degree of coincidence of the voice signals to the fixed time determination unit 12.

  Note that the degree of coincidence of audio signals is a numerical value indicating how much the input audio signal and the reference audio signal match. Moreover, it becomes a standard for determining whether the destination floor has been recognized correctly or incorrectly by the degree of coincidence of the audio signals.

  The confirmation time determination means 12 determines the optimum confirmation time by using the time until the passenger's destination floor recognized by the voice recognition means 11 is called and registered by the call registration processing means 13 as the confirmation time, and the call registration processing means 13 To enter. This optimal decision time is determined as follows.

  That is, the fixed time determination means 12 stores in advance a fixed time table in which the fixed time and the coincidence degree of the voice signal are associated with each other in a storage unit (not shown), and the voice recognition means is selected from the fixed time table. The optimum decision time is determined by extracting the decision time corresponding to the degree of coincidence of the audio signals calculated by 11.

  Note that this determination time table may be set such that the higher the matching degree, the shorter the corresponding determination time, and the lower the matching degree, the longer the corresponding determination time.

  The call registration processing unit 13 sets the destination floor that the voice recognition unit 11 determines that call registration is possible when the optimum fixed time determined by the fixed time determination unit 12 has passed as the destination floor to be registered for the elevator control panel 30. To register the destination floor call.

  The elevator control panel 30 controls the operation of the elevator in order to carry passengers to the destination floor in response to the call registration of the destination floor.

  Next, the details of the speech recognition processing executed by the call registration device 10 will be described with reference to the flowchart of FIG.

  FIG. 2 is a flowchart showing the speech recognition process in Embodiment 1 of the present invention. First, the call registration device 10 receives an audio signal input from the audio input device 20 in step S201.

  Next, in step S202, the call registration device 10 recognizes the passenger's destination floor from the input audio signal. In step S203, the call registration device 10 compares the input audio signal with the reference audio signal corresponding to each destination floor stored in advance, and calculates the degree of coincidence of these audio signals.

  Next, in step S204, the call registration device 10 determines whether the recognized destination floor is call-registerable according to the calculated degree of coincidence of the audio signals. In step S204, the call registration device 10 ends the series of processes when determining that the recognized destination floor is not call-registerable because the calculated degree of coincidence of the audio signals is less than the specified value.

  On the other hand, if the call registration device 10 determines in step S204 that the calculated coincidence of the audio signals is equal to or greater than the specified value, and determines that the recognized destination floor can be registered as a call, in step S205, The optimum decision time is determined by extracting the decision time corresponding to the calculated coincidence of the audio signals from the decision time table stored in advance.

  Next, in step S206, the call registration device 10 determines again whether another audio signal is input from the audio input device 20 during the elapse of the determined optimum determination time.

  If the call registration device 10 determines in step S206 that another audio signal has been input, the call registration device 10 executes the processing in step S202 and subsequent steps again.

  On the other hand, if the call registration device 10 determines in step S206 that another audio signal has not been input, the destination floor that is determined to be call-registerable in step S207 is determined as the destination floor to be registered. To do.

  Next, in step S208, the call registration device 10 outputs the confirmed destination floor call to the elevator control panel 30 to register the destination floor call. Finally, a series of processing ends.

  The elevator control panel 30 controls the operation of the elevator in order to carry passengers to the destination floor in response to the call registration of the destination floor.

  As a result, when the degree of coincidence of the voice signal is low (that is, when the destination floor is considered to have been recognized by voice), the lower the degree of coincidence, the longer the decision time, so that the passenger speaks again. The call registration can be changed.

  In addition, when the degree of coincidence of voice signals is high (that is, when the destination floor is considered to have been correctly recognized by voice), the higher the degree of coincidence, the shorter the confirmation time. There is little waiting time until it is done.

  As described above, according to the first embodiment of the present invention, the call registration device stores in advance, with the input voice signal, the fixed time from the voice recognition of the passenger's destination floor to the call registration. It can be determined according to the degree of coincidence with the reference audio signal. Thereby, when the degree of coincidence is low (it is considered that the destination floor is erroneously recognized by voice), the determination time becomes long. Therefore, the passenger can change the call registration by emitting a voice again. On the other hand, when the degree of coincidence is high (it is considered that the destination floor is correctly recognized by voice), the determination time is shortened. For this reason, it is possible to eliminate an unnecessary waiting time until the call is registered after the voice is recognized.

Embodiment 2. FIG.
In the first embodiment, the case where the call registration device 10 determines the optimum determination time according to the degree of coincidence of audio signals has been described. On the other hand, in the second embodiment of the present invention, a case will be described in which the optimum fixed time is determined in accordance with the operation state of the elevator input from the elevator control panel 30 in addition to the degree of coincidence of the audio signals. .

  Here, in the elevator call registration device according to the second embodiment, the time required for the passenger's destination floor to concentrate on a specific floor in addition to the degree of coincidence of the audio signals as compared with the first embodiment. Considering the case where the elevator is operating in the belt, it has the technical feature of determining the optimum fixed time.

  Examples of the time zone when the passenger's destination floor is concentrated on a specific floor include, for example, the lunch time zone when the passenger's destination floor is concentrated on the cafeteria floor, the return time zone when the passenger's destination floor is concentrated on the entrance floor, and the passenger's destination There are office hours when the floor is concentrated on the office floor, and meeting hours when the passenger's destination floor is concentrated on the conference floor.

  Therefore, in order to specifically describe the call registration device 10 according to the second embodiment, in a lunch time zone where the passenger's destination floor is concentrated on the cafeteria floor (hereinafter referred to as a time zone where the lunchtime service is applied). The details of the speech recognition process executed by the call registration device 10 when the elevator is operating will be described with reference to the flowchart of FIG.

  FIG. 3 is a flowchart showing the speech recognition process in Embodiment 2 of the present invention. First, the call registration device 10 sequentially executes the processes from step S201 to step S204 similar to those in the first embodiment.

  When the call registration device 10 determines in step S204 that the calculated coincidence of the audio signals is less than the specified value and thus the recognized destination floor cannot be registered as a call, the first embodiment 1 described above. A series of processing is ended in the same manner as in the above.

  On the other hand, when the degree of coincidence of the calculated audio signal is equal to or greater than the specified value in step S204, if it is determined that the recognized destination floor can be registered for call registration, in step S301, the operation state of the elevator is confirmed. Next, it is determined whether or not it is a time period when the lunchtime service is applied.

  If the call registration device 10 determines in step S301 that it is not the time zone to which the lunchtime service is applied, the call registration device 10 sequentially executes the processing from step S205 to step S208 similar to the first embodiment. Then, a series of processing is completed. In step S205, the fixed time table in the first embodiment is used.

  On the other hand, if the call registration device 10 determines in step S301 that the lunch time service is being applied, the call registration device 10 determines in step S302 the degree of coincidence of the calculated audio signal and the confirmed elevator operating state. Accordingly, the optimum decision time is determined as follows.

  That is, the call registration device 10 according to the second embodiment stores an individual fixed time table corresponding to each time zone, and in each fixed time table, the fixed time and the matching degree of the audio signal are associated with each other. Yes.

  And the call registration apparatus 10 selects the fixed time table corresponding to the time slot | zone to which a lunch time service is applied among the fixed time tables corresponding to each time slot | zone in this case. Then, an optimum decision time is determined by extracting a decision time corresponding to the calculated degree of coincidence of the audio signals from the decision time table.

  In the second embodiment, as in the first embodiment, the higher the matching degree, the shorter the corresponding fixed time, and the lower the matching degree, the longer the corresponding fixed time. Should be set.

  Furthermore, in this Embodiment 2, compared with previous Embodiment 1, even if it is the same audio | voice signal coincidence, it is a time slot | zone to which a lunchtime service is applied, and a destination floor is a cafeteria floor. In this case, the fixed time table may be set so that the fixed time corresponding to the degree of coincidence is further shortened.

  Finally, the call registration device 10 sequentially executes the processes in steps S206 to S208 similar to those in the first embodiment, and then ends the series of processes.

  As a result, compared with the first embodiment, even when the degree of coincidence is the same, it is a time zone in which the lunchtime service is applied, and when the destination floor is the dining hall floor, the confirmation time is further shortened. Become. For this reason, passengers who go to the cafeteria floor during the time period when the lunchtime service is applied have less waiting time from voice recognition to call registration.

  As described above, according to the second embodiment of the present invention, the call registration device determines the coincidence of the audio signal from the fixed time table corresponding to the time zone in which the passenger's destination floor is concentrated on a specific floor. The corresponding fixed time can be extracted. When the destination floor is a specific floor, the fixed time can be further shortened. This makes it possible to obtain an elevator call registration device that can further eliminate unnecessary waiting time from voice recognition to call registration.

  In addition, the operation state of the elevator which the call registration apparatus 10 in this Embodiment 2 confirms is not restricted to what was illustrated here. In other words, the call registration device 10 can be confirmed individually in other time zones (returning time zone, leaving time zone, meeting time zone, etc.) in which the passenger's destination floor is concentrated on a specific floor in the elevator operation state. The same effect can be obtained by using the time table.

Embodiment 3 FIG.
In the second embodiment, the operation of the registration device 10 in the case where the elevator is operating in a time zone in which the passenger's destination floor is concentrated on a specific floor has been described. In contrast, in the third embodiment of the present invention, a call registration device 10 that determines a fixed time according to a floor to which an audio signal is input will be described.

  Here, the elevator call registration device according to the third embodiment determines an optimum determination time in consideration of the floor to which the audio signal is input as compared with the second embodiment. Has technical features. Instead of considering the time zone as in the second embodiment, it is possible to consider only the floor to which the audio signal is input.

  Here, when determining the optimum confirmation time, the floor to which the audio signal is input is also considered. For example, the audio signal is input from the floor where many people are accustomed to call registration (voice registration). And so on. Since passengers from such floors are unlikely to change the call registration, the confirmation time can be shortened.

  Here, the floor to which the audio signal is input is a floor corresponding to the position of the elevator car when the voice recognition process performed by the passenger is performed.

  Therefore, in order to specifically describe the call registration device 10 according to the third embodiment, the elevator is used in a time zone in which the passenger's destination floor is concentrated on the office floor (hereinafter referred to as a time zone in which the up-peak service is applied). FIG. 4 shows the details of the voice recognition processing executed by the call registration device 10 in response to a voice signal input from the office floor where there are many people who are used to voice registration. This will be described with reference to the flowchart of FIG.

  FIG. 4 is a flowchart showing voice recognition processing in Embodiment 3 of the present invention. First, the call registration device 10 sequentially executes the processes from step S201 to step S204 similar to those in the first embodiment.

  If the call registration device 10 determines in step S204 that the calculated coincidence of the audio signals is less than the specified value and thus the recognized destination floor is not call-registerable, it is the same as in the first embodiment. A series of processing is finished.

  On the other hand, in step S204, since the calculated degree of coincidence of the audio signals is equal to or greater than the specified value, if it is determined that the recognized destination floor can be called and registered, the operation state of the elevator is confirmed in step S401. Next, it is determined whether or not the up-peak service is applied.

  If the call registration device 10 determines in step S401 that it is not the time period in which the up-peak service is applied, the call registration device 10 sequentially executes the processes in steps S205 to S208 similar to those in the first embodiment. Then, a series of processing is completed. In step S205, the fixed time table in the first embodiment is used.

  On the other hand, if the call registration device 10 determines in step S401 that the up-peak service is being applied, in step S402, passengers get on from the office floor where many people are accustomed to voice registration. Therefore, it is determined whether or not the floor to which the audio signal is input is an office floor.

  If the call registration device 10 determines in step S402 that the floor to which the audio signal is input is an office floor, the call registration device 10 determines the degree of coincidence of the calculated audio signal and the confirmed operation of the elevator in step S403. Depending on the state, the optimum decision time is determined as follows.

  That is, the call registration device 10 according to the third embodiment further stores an individual fixed time table corresponding to each floor to which the audio signal is input, and each fixed time table includes the fixed time and the audio signal. Is associated with the degree of match.

  And the call registration apparatus 10 selects the fixed time table corresponding to the case where the floor to which the audio | voice signal was input is an office floor among those fixed time tables. Then, an optimum decision time is determined by extracting a decision time corresponding to the calculated degree of coincidence of the audio signals from the decision time table.

  In the third embodiment, as in the previous first embodiment, the higher the matching degree, the shorter the corresponding determination time, and the lower the matching degree, the longer the corresponding determination time. Should be set.

  Furthermore, in the third embodiment, when the floor to which the voice signal is input is an office floor where many people are accustomed to voice registration, the same voice signal matches as compared with the first embodiment. Even if it is a degree, an individual fixed time table corresponding to each floor may be set so that the fixed time corresponding to the degree of coincidence is further shortened.

  And the call registration apparatus 10 complete | finishes a series of processes, after performing sequentially the process of step S206-step S208 similar to previous Embodiment 1. FIG.

  On the other hand, if the call registration device 10 determines in step S402 that the floor to which the audio signal has been input is not an office floor, in step S404, the coincidence of the calculated audio signal and the confirmed elevator The optimum fixed time can be determined in the same manner as in the second embodiment in accordance with the operation state.

  That is, the call registration device 10 selects a fixed time table corresponding to a time zone in which the up-peak service is applied. Then, the optimum decision time can be determined by extracting the decision time corresponding to the calculated coincidence of the audio signals from the decision time table.

  Finally, the call registration device 10 sequentially executes the processes in steps S206 to S208 similar to those in the first embodiment, and then ends the series of processes.

  Note that the voice recognition process executed by the call registration device 10 in the third embodiment is the same as that in the second embodiment (step S205) when the up-peak service is not applied. Executed. However, even if the up-peak service is not applied, if the floor to which the audio signal is input is an office floor, the same processing as in step S403 (that is, an individual fixed time corresponding to each floor) The process may be changed to execute a process of extracting a fixed time using a table.

  As a result, the passenger is in the time zone where the up-peak service is applied even if the degree of coincidence is the same as in the first embodiment, and the destination floor is the office floor or from the office floor. When a voice signal is input, the determination time is further shortened, so that the waiting time from voice recognition to call registration is reduced.

  As described above, according to the third embodiment of the present invention, the call registration device determines a decision corresponding to the degree of coincidence of the audio signal from the individual decision time table corresponding to each floor to which the audio signal is input. Time can be extracted. In addition, when an audio signal is input from a specific floor, the determination time can be further shortened. This makes it possible to obtain an elevator call registration device that can eliminate unnecessary waiting time from voice recognition to call registration.

  In addition, the operation state of the elevator which the call registration apparatus 10 in this Embodiment 3 confirms is not restricted to what was illustrated here. That is, the call registration device 10 can obtain the same effect even when the specific floor to which the audio signal is input is not the office floor but another floor in the elevator operation state.

  DESCRIPTION OF SYMBOLS 10 Call registration apparatus, 11 Voice recognition means, 12 Fixed time determination means, 13 Call registration process means, 20 Voice input device, 30 Elevator control panel.

Claims (4)

An elevator call registration device that is connected to an elevator control panel that controls the operation of the elevator according to the destination floor registered for call, and that performs call registration for the destination floor with respect to the elevator control panel based on voices made by passengers. There,
Reference audio signals corresponding to each destination floor are stored in advance, the destination floor is recognized from an audio signal converted from a voice emitted by a passenger, and the audio signal is compared by comparing the audio signal with the reference audio signal. Voice recognition means for calculating a degree of coincidence and determining that the recognized destination floor can be called and registered according to the calculated degree of coincidence;
A fixed time table associating the fixed time, which is the time from the recognition of the destination floor determined to be call registration by the voice recognition means to the call registration of the destination floor, with the degree of coincidence is stored in advance. A fixed time determining means for extracting a fixed time corresponding to the degree of coincidence calculated by the voice recognition means from the fixed time table;
When the fixed time extracted by the fixed time determining means has passed, the destination floor determined to be call-registerable by the voice recognition means is output to the elevator control panel as a destination floor to be call-registered, An elevator call registration device comprising: call registration processing means for performing call registration on the destination floor. In the elevator call registration device according to claim 1,
The fixed time determination means has an individual fixed time table for each time zone as the fixed time table stored in advance, and the fixed time corresponding to the degree of coincidence calculated by the voice recognition means An elevator call registration device, characterized in that it is extracted from the individual fixed time table corresponding to a band. In the elevator call registration device according to claim 1 or 2,
The fixed time determining means has a fixed fixed time table for each floor corresponding to a car position when performing recognition processing of a voice emitted by a passenger as the fixed time table stored in advance. An elevator call registration device characterized in that a fixed time corresponding to the degree of coincidence calculated by the voice recognition means is extracted from the individual fixed time table corresponding to the floor. An elevator call registration device that is connected to an elevator control panel that controls the operation of the elevator according to the destination floor registered for call, and that performs call registration for the destination floor with respect to the elevator control panel based on voices made by passengers. An elevator call registration method to be executed,
A first storage step of previously storing a reference audio signal corresponding to each destination floor in the storage unit;
Recognizing the destination floor from the voice signal converted from the voice emitted by the passenger, calculating the degree of coincidence of the voice signal by comparing the voice signal and the reference voice signal stored in the storage unit, the calculated match Depending on the degree, a speech recognition step of determining that the recognized destination floor can be called and registered,
A storage unit that stores a fixed time table associating a fixed time that is a time from recognition of the destination floor, which is determined to be call registration possible in the voice recognition step, to call registration of the destination floor, and the degree of coincidence; A second storage step to be stored in advance,
A fixed time determination step of extracting a fixed time corresponding to the degree of coincidence calculated in the voice recognition step from the fixed time table stored in the storage unit;
When the fixed time extracted in the fixed time determination step elapses, the destination floor determined to be call-registerable by the voice recognition step is output to the elevator control panel as a destination floor to be call-registered, An elevator call registration method comprising: a call registration processing step for performing call registration on the destination floor.

Images