JP5402219B2 - Elevator call registration device - Google Patents

Description

  The present invention relates to an elevator call registration device.

In conventional elevator call registration devices, call registration is performed by recognizing voices emitted by elevator users, and installed in elevator halls to prevent misrecognition due to noise from other users. A microphone, a person detecting means for detecting whether or not an elevator user is standing in front of the microphone, and an audio signal input to the microphone only when there is a person detection signal from the person detecting means. There is known a voice recognition means that generates a call registration signal when a voice signal matches a registered voice as compared with a voice registered in advance (see, for example, Patent Document 1).
Also, by providing a similar call registration device in the car, car destination floor call registration can be performed by voice recognition.

Japanese Patent No. 2557939

  When the conventional elevator call registration device shown in Patent Document 1 is provided in the car and the car destination floor call registration is performed by voice recognition, the car is in a relatively narrow space that is almost sealed. For this reason, it is necessary to consider the influence of the reflection of the sound emitted by the elevator user, that is, the acoustic model in the car.

  However, the conventional elevator call registration device disclosed in Patent Document 1 does not consider the acoustic model in the car, and reverberation sound is more superimposed in the door-closed state than in the door-open state. The floor that the user speaks due to the effects of changes in the reverberant sound in the car due to the door opening and closing state, the effects of changes in the sound reflection due to differences in the car material, and the influence of the running sound of the car There is a problem in that the floor information (for example, “first floor” or the like) cannot be accurately recognized by voice, and erroneous recognition may occur.

  The present invention has been made to solve such a problem, and an error caused by a change in an elevator state (door open / closed state, traveling stop state, etc.) in speech recognition using a microphone provided in a car. It is possible to obtain an elevator call registration device that can prevent recognition and can accurately recognize speech.

In the elevator call registration device according to the present invention, a microphone that is provided in an elevator car and receives a voice input by a user and outputs it as a voice signal, and an elevator that outputs state information relating to the state of the elevator. Of the control device, the audio signal output from the microphone, and the registered audio signal that is audio information of a destination floor issued by a user registered in advance for each state of the elevator, the corresponding to the state information Voice registration means for comparing the registered voice signal in the state of the elevator and outputting a call registration signal when they match , the status information includes door opening / closing information of the car door of the car and The elevator state corresponding to the state information includes at least the traveling information of the car, Defined by closed information and combinations of the travel in information, door open state, during door opening and closing operation, and at least including up each state during running in the stop and door-closed state in the door-closing state.

Call registration device for an elevator according to the invention, in the speech recognition using a microphone provided in the car, it is possible to prevent erroneous recognition due to the change of state of the elevator, can accurately speech recognition There is an effect that there is.

It is a block diagram which shows the whole structure of the elevator call registration apparatus which concerns on Embodiment 1 of this invention. It is a table | surface which shows the specific example of the situation judgment result produced by the situation judgment means which concerns on Embodiment 1 of this invention. It is explanatory drawing which shows the memory position in the address space of the registration audio | voice information of each environment memorize | stored in the memory | storage means which concerns on Embodiment 1 of this invention.

Embodiment 1 FIG.
FIGS. 1 to 3 relate to Embodiment 1 of the present invention, FIG. 1 is a block diagram showing the overall configuration of an elevator call registration device, and FIG. 2 is a concrete example of a situation judgment result created by situation judgment means. FIG. 3 is an explanatory diagram showing storage positions in the address space of registered voice information under each environment stored in the storage means.

In FIG. 1, reference numeral 1 denotes a car that is disposed in a hoistway (not shown) of the elevator so as to be lifted and lowered. The elevator operation including the raising and lowering of the car 1 is controlled by the elevator control device 2.
In the car 1, a destination floor registration button 3 for registering a car destination floor call by a user's operation is installed corresponding to each destination floor. The operation of the destination floor registration button 3 is performed. The situation is transmitted to the elevator control device 2, and the elevator control device 2 gives an instruction to turn on the button lamp of the destination floor registration button 3.

A microphone 4 is also installed in the car 1 for receiving voice input from the user, converting the voice into an electrical signal, and outputting the electrical signal. The microphone input by the user is input through the microphone 4. The voice information is recognized by the voice recognition means 5.
This voice recognition means 5 is a state judgment means 6 for judging the state of the elevator (door open / closed state, running stop state, etc.), and a storage means for preliminarily storing a registered voice signal which is voice information of a destination floor issued by the user ( (Storage unit) 7, based on the situation determination result by the state determination unit 6, the registered audio signal selection unit 8 that selects an appropriate audio signal from the registered audio signal stored in the storage unit 7, and the microphone 4 A voice comparison means 9 is provided for comparing the voice signal with the registered voice signal selected by the registered voice signal selection means 8.

The elevator control device 2 changes the state of the elevator every time the state of the elevator changes, that is, door opening / closing information 10 that is information on the opening / closing state of a car door (not shown) provided in the car 1, and the elevator, In-travel information 11, which is information regarding whether or not the car 1 is traveling, is output to the state determination means 6.
Here, the door opening / closing information 10 takes three states according to the state of the car door: an open state, a closed state, and a door opening / closing operation, and more specifically, “1” in the open state and “1” in the closed state. Each value of “3” is taken during the door opening / closing operation. Further, the traveling information 11 takes two states of traveling and stopping according to the state of the car 1, more specifically, “1” when the car 1 is traveling and “2” when the car 1 is stopped. Each value.

The state determination unit 6 determines the current state of the elevator based on the door opening / closing information 10 and the traveling information 11 from the elevator control device 2, and this result is used as the state determination result 12 to the registered voice signal selection unit 8. Is output.
The state determination here is performed according to the rules of the table shown in FIG. 2 (in this table, “*” is “Don't care”, that is, the value of the information does not affect the state determination. Meaning that it does n’t.)

  That is, when the value of the door opening / closing information 10 is “1” (open state), the state determination result 12 is a value “1” because the elevator state is the door open state regardless of the value of the traveling information 11. Take. When the value of the door opening / closing information 10 is “3” (opening / closing operation), the state determination result 12 is “2”, assuming that the elevator is in the door opening / closing operation regardless of the value of the traveling information 11. Take. When the value of the door opening / closing information 10 is “2” (closed state) and the value of the traveling information 11 is “2”, the state of the elevator is assumed to be the door open state (stopped). The determination result 12 takes a value of “3”. When the value of the door opening / closing information 10 is “2” (closed state) and the value of the traveling information 11 is “1”, the elevator state is the door open state (traveling). As a result, the state determination result 12 takes a value of “3”.

The storage means 7 stores in advance a registered voice signal that is voice information of the destination floor spoken by the user in each of the four elevator states that the state determination result 12 can take.
An example of the method of storing the registered audio signal in the storage means 7 is shown in FIG. Here, for example, among the four situations that the state judgment result 12 can take, the storage area of the registered audio signal in one situation is 0x0400 bytes (the leading “0x” is expressed in hexadecimal notation) It is assumed that it is secured).

  Therefore, in this example, the registered voice signal when the elevator is in the door open state (the value of the state determination result 12 is “1”) is stored in an area of 0x0000 to 0x0399 bytes in the address space, and the door opening / closing operation is being performed ( The registered voice signal when the value of the state determination result 12 is “2”) is stored in an area of 0x0400 to 0x0799 bytes in the address space, and in the door closed state (stopped) (the value of the state determination result 12 is “3”). The registered voice signal is stored in an area of 0x0800 to 0x0B99 bytes in the address space, and the registered voice signal in the door closed state (running) (the value of the state determination result 12 is “4”) is 0x0C00 in the address space. It is stored in an area of 0x0F99 bytes.

Based on the state determination result 12 from the state determination unit 6, the registered sound signal selection unit 8 selects and acquires the registration sound signal corresponding to the state determination result 12 from the storage unit 7. Is output to the voice comparison means 9. The selection and acquisition method here will be described in detail with reference to the above example.
That is, the registered voice signal selection means 8 calculates the head address where the registered voice signal corresponding to the state determination result 12 is stored by the following equation (1).

  ((Value of state determination result 12) -1) × (0x0400) (1)

For example, when the value of the state determination result 12 is “1”, that is, the door is open, the calculation result by the equation (1) is 0x0000, and this registered sound signal corresponding to the value “1” of the state determination result 12 is 0x0000. When it is determined that the stored address is the head address and the value of the state determination result 12 is “3”, that is, the door is closed (stopped), the calculation result by the expression (1) is 0x0800, and 0x0800 is the state determination result. It can be seen that this is the head address where the registered audio signal corresponding to the value “3” of 12 is stored.
In this way, the registered voice signal selection means 8 selects and acquires the corresponding registered voice signal from the storage means 7 based on the state determination result 12 from the state determination means 6, and obtains this as the registered voice signal 13 as the voice comparison means. Output to 9.

  On the other hand, the voice information of the destination floor uttered by the user and inputted to the microphone 4 in the car 1 is converted into a digital voice signal 14 by an A / D converter (analog / digital converter) provided in the microphone 4. Is output to the voice comparison means 9.

The voice comparison means 9 compares the digital voice signal 14 from the microphone 4 with the registered voice signal 13 from the registered voice signal selection means 8 and determines whether or not these voice signals match. It is determined whether or not the voice information represented by the signal 14 matches the elevator movement represented by the registered voice signal 13, that is, the voice information related to the destination floor.
When the audio information represented by the digital audio signal 14 matches the audio information related to the destination floor represented by the registered audio signal 13, the information on the destination floor is output to the elevator control device 2 as the call registration signal 15. .

The elevator control device 2 that has received the call registration signal 15 from the voice comparison means 9 registers the car destination floor call with the destination floor as the destination floor based on the destination floor information included in the call registration signal 15. And instructing to turn on the button lamp of the destination floor registration button 3 corresponding to the destination floor.
Then, the button light of the corresponding destination floor registration button 3 is turned on, and the user can know that the car destination floor call to the destination floor spoken by himself / herself is correctly registered.

  Here, as the state information used to determine the state of the elevator, the door opening / closing information and the traveling information are used here, but this state information is limited to the door opening / closing information and the traveling information. It is also possible to use information on the state of the elevator that may affect other voice recognition.

  In the elevator call registration device configured as described above, a microphone that receives a voice input from a user and outputs it as a voice signal is provided in the elevator car, and the elevator control device relates to the state of the elevator. Elevator control out of the voice signal output from the microphone by the voice recognition means and the registered voice signal which is the voice information of the destination floor issued by the user registered in advance for each state of the elevator. Speech recognition using a microphone provided in the car by comparing the registered voice signal in the state of the elevator corresponding to the status information from the device and outputting a call registration signal when they match Can prevent misrecognition due to changes in the state of the elevator, and accurately recognize voice. It is possible.

  At this time, in view of the fact that changes in the reverberant sound in the car due to the open / closed state of the door and the running sound of the car are considered to have a significant effect on voice recognition, the door information of the car car door is opened and closed as status information. It includes at least information and running information of the car, and the state of the elevator corresponding to the status information is determined by a combination of the door opening and closing information and the running information, and is in a door open state and a door opening and closing operation. Improve the accuracy of voice recognition by including at least each state of stopping in the door closed state (door closed state (stopped)) and running in the door closed state (door closed state (running)). Can be planned.

DESCRIPTION OF SYMBOLS 1 Passenger car 2 Elevator control device 3 Destination floor registration button 4 Microphone 5 Voice recognition means 6 State judgment means 7 Storage means 8 Registration voice signal selection means 9 Voice comparison means 10 Door open / close information 11 Running information 12 State judgment result 13 Registered voice Signal 14 Digital audio signal 15 Call registration signal

Claims (2)

A microphone that is provided in the elevator car and receives a voice input from a user and outputs a voice signal;
An elevator control device that outputs state information relating to the state of the elevator;
In the state of the elevator corresponding to the state information, the sound signal output from the microphone and the registered sound signal which is the sound information of the destination floor issued by the user registered in advance for each state of the elevator Voice recognition means for comparing the registered voice signal and outputting a call registration signal when they match ,
The status information includes at least door opening / closing information of the car door of the car and running information of the car,
The state of the elevator corresponding to the state information is determined by a combination of the door opening / closing information and the traveling information, and is in a door open state, a door opening / closing operation, a door closed state, and a door closed state. An elevator call registration device comprising at least each of the following states: The voice recognition means
State determination means for determining the state of the elevator based on the state information output from the elevator control device and outputting a state determination result;
Storage means for storing in advance for each state of the elevator, a registered voice signal which is voice information of a destination floor issued by a user;
The registered sound signal in the state of the elevator corresponding to the state determination result output from the state determining means is selected from the registered sound signals stored in the storage means, and the selected registered sound signal is output. Registered audio signal selection means to
Comparing the audio signal output from the microphone with the registered audio signal output from the registered audio signal selection unit, and a voice comparison unit that outputs a call registration signal when they match ,
Said storage means includes at least claim 1, wherein door open state, during door opening and closing operation, that stores in advance the registered voice signal for each state of traveling in stop and door-closed state in door-closed state The elevator call registration device described in 1.

Images