JP6672722B2 - Voice control device for vehicles - Google Patents

Description

  The present disclosure relates to a voice operation device for operating various on-board operation devices mounted on a vehicle by voice.

  2. Description of the Related Art A voice operation device for a vehicle has been developed which operates various devices (on-vehicle operation device) mounted on a vehicle by a voice of a passenger of the vehicle such as a driver. In a conventional vehicle voice operation device, in order to prevent the driver from being distracted by voice input during driving, the input of voice operation is restricted in accordance with the driving condition of the vehicle, thereby improving driver safety. We support driving. For example, the vehicle voice operation device disclosed in Patent Literature 1 determines the degree of danger that the vehicle is parked through the determination of the traveling state of the vehicle, and performs voice operation on the operation target device according to the determined degree of danger. By limiting the input, more accurate safe driving support is achieved.

JP-A-2004-217101

  As described above, the conventional vehicle voice operation device may limit the input of the voice operation according to the driving situation, but when the voice operation is input, the operation of the in-vehicle operation device is performed according to the voice operation. Execute. However, depending on the situation where the vehicle is placed, if the operation is performed in accordance with the voice operation of the speaker, there is a possibility that safe driving may be hindered. For example, if a voice operation such as turning off the headlights during night driving or running in a tunnel is performed, and this operation is performed as it is, it may be an obstacle to safe driving. In addition, the operation content of the in-vehicle operation device (operation target device) to be operated can be estimated from the content of the utterance of the driver, etc., and when the appropriate operation is performed, the burden on the driver to re-execute the operation instruction is reduced. It is expected that a comfortable driving environment can be provided to the driver.

  In view of the above circumstances, at least one embodiment of the present invention aims to provide a vehicle voice operation device that controls execution of a voice operation in light of a state where a vehicle is parked.

(1) A voice operation device for a vehicle according to at least one embodiment of the present invention includes:
A vehicle state information acquisition unit that acquires vehicle state information indicating a state of the vehicle,
An input voice recognition unit that converts input voice into voice recognition data,
On the basis of the voice recognition data, an in-vehicle operation device serving as an operation target device and a voice recognition data analysis unit that determines the operation content for the operation target device,
An operation execution determination unit that determines the validity of executing the operation content on the operation target device based on the vehicle state information,
An operation instruction generation unit that generates an operation instruction for executing the operation content on the operation target device in response to an instruction from the operation execution determination unit;
A device operation unit that executes the operation command;
An output sound generation unit that converts the notification data received from the operation execution determination unit into a sound,
The operation execution determination unit does not execute the operation command when determining that there is no validity, and outputs the notification data for outputting a response message indicating that there is no validity in the output voice. Output to the generator.

  According to the configuration of the above (1), the vehicle voice operation device captures the voice (utterance) of the driver of the vehicle, etc., and performs the voice operation contents (operation target device and operation content) intended by the speaker by voice recognition. ) Is determined (guessed). Then, the vehicle voice operation device executes a voice operation (operation command C) for reflecting the voice operation on the vehicle in light of the current state (vehicle state information) where the vehicle is placed. Is determined. Then, when it is determined that the execution of the voice operation is not appropriate, the driver is notified of the fact without performing the voice operation. By controlling the execution of the voice operation in accordance with the state of the vehicle in this way, for example, it is possible to prevent a dangerous state from occurring, such as performing a voice operation such as turning off a headlight while traveling in a tunnel, It can support safe driving.

(2) In some embodiments, in the configuration of the above (1),
The vehicle has a sensor for detecting at least one of a driving state of the vehicle or peripheral information of the vehicle as vehicle-side information,
The vehicle state information includes the vehicle side information,
The operation execution determination unit, based on the vehicle state information including the vehicle-side information, to determine the validity of executing the operation content to the operation target device,
If the execution of the operation command for the operation target device is determined to be incompatible with the vehicle-side information associated with the operation target device in light of the function of the operation target device, it is determined that there is no validity. If it is determined that the information matches the vehicle-side information associated with the operation target device, it is determined that there is validity.
According to the configuration of the above (2), the vehicle state information includes at least the vehicle side information, and the content of the voice operation of the in-vehicle operation device is the content of the vehicle related to the operation target device obtained by the sensor installed in the vehicle. The validity is determined by comparison with the state. As a result, for example, it is possible to prevent execution of an operation command to turn off the headlights when the vehicle is traveling in the tunnel, and it is possible to support safe driving.

(3) In some embodiments, in the configuration of the above (1),
The vehicle has a communication device for communicating with the outside of the vehicle,
The vehicle state information includes external information obtained via the communication device,
The operation execution determination unit, based on the vehicle state information including the external information, to determine the validity of executing the operation content to the operation target device,
Based on the external information, within a predetermined time after the execution of the operation command for the operation target device, a prediction is made that the operation command for the operation target device is to be executed on the operation target device in a direction opposite to the operation command for the operation target device. In this case, it is determined that there is no validity, and when the prediction does not hold, it is determined that there is validity.

  According to the configuration of (3), external information such as information on the surroundings viewed from the position information of the vehicle, information on traffic congestion around the route to the destination, and construction information are acquired using the communication device. Also, the validity of the voice operation of the in-vehicle operation device is determined by comparing it with external information. For example, when there is a voice operation to the effect of the wiper operation, if it is predicted from external information that the vehicle will enter the tunnel immediately after execution of the voice operation, even if the wiper is operated immediately before the tunnel, It is expected that a voice operation to stop the wiper will be performed by immediately entering the tunnel. As described above, even when the first operation command is executed in accordance with the intention of the speaker, if it is predicted that the second operation will immediately execute the opposite operation command, the vehicular voice operation device may be used. Is configured not to immediately execute an operation command for the operation target device. Thus, it is possible to prevent a situation in which the speaker repeats the voice operation, and to support a safe driving by forming a situation in which the driver can concentrate on driving.

(4) In some embodiments, in the above configurations (1) to (3),
The vehicle state information includes operating state information of the vehicle-mounted operation device,
The operation execution determination unit, based on the vehicle state information including the operation state information of the operation target device, to determine the validity of executing the operation content to the operation target device,
When it is determined that the operation state information of the operation target device and the operation content match, it is determined that there is no validity, and the operation state information and the operation content of the operation target device are When it is determined that they do not match, it is determined that the validity exists.

  According to the above configuration (4), the validity of the voice operation of the in-vehicle operation device is determined by being compared with the operation state of the in-vehicle operation device. As a result, for example, in a state where the wiper is operating, it is possible to prevent an unnecessary load such as executing an operation command for operating the wiper from being applied to the device operation unit, and the response (responsiveness) of the vehicle voice operation device is prevented. ) Can be increased.

(5) In some embodiments, in the configuration of the above (4),
In the case where the vehicle state information includes, in addition to the operation state information, other information for determining the validity of executing the operation content on the operation target device, the operation target device As a result of determining the validity based on the operation state information, the validity is further determined based on the other information only when the validity is determined.
According to the configuration of the above (5), when it is determined that the voice operation based on the operation state information of the operation target device is valid, the voice operation based on other information such as the vehicle side information or the external information is performed. A validity determination is performed. Conversely, when it is determined that the voice operation based on the operation state information is not appropriate, the validity of the voice operation based on other information (e.g., vehicle-side information or external information) is not determined. It is determined that the voice operation is not valid. To determine the validity of the voice operation based on the operation state information of the operation target device, it is only necessary to confirm the operation state of the operation target device. Since it does not depend on the sense or situation, it can be performed more accurately than other information. Therefore, the accuracy of interactive communication with the speaker can be increased, the response of the vehicle voice operation device can be improved, and the reliability of the vehicle voice operation device can be improved.

(6) In some embodiments, in the above configurations (4) to (5),
The in-vehicle operating device is configured to be able to set an operation state in stages,
The operating state information includes setting step information indicating a setting step of the operating state,
The operation execution determination unit, when the operation content for the operation target device is determined to be an operation for relatively changing the setting stage of the operation state, the vehicle state information acquisition unit acquires the The validity of executing the operation content on the operation target device is determined based on the vehicle state information including the setting stage information.
According to the configuration of the above (6), the voice operation such as changing the operation stage of the operation target device such as the strength of the wiper is compared with the state of the vehicle obtained by a sensor or the like installed in the vehicle, Its validity is determined. For example, when it is determined that it is appropriate to maximize the operation level of the wiper due to heavy rain legs, and the content of the voice operation does not conform to this determination, it is determined that the above-described validity is not obtained. Thus, the in-vehicle operation device can be guided to an appropriate operation state according to the traveling state of the vehicle, and safe driving can be supported.

(7) In some embodiments, in the above configurations (1) to (6),
The notification data, together with a response message indicating that there is no validity, includes a question message regarding the execution of the operation content,
When receiving a response to the question message via the input voice recognition unit, the voice recognition data analysis unit executes the instruction to the operation command generation unit according to the content of the response, or And does not execute the operation content on the operation target device.

  According to the configuration of the above (7), when the above-mentioned validity is not obtained, the question message is notified, and the execution of the voice operation is controlled in accordance with the response of the speaker, thereby confirming the intention of the speaker regarding the voice operation. . As a result, the accuracy of the interactive exchange with the speaker can be increased, and the reliability of the vehicle voice operation device can be increased.

(8) In some embodiments, in the configurations of the above (1) to (7),
A voice input device into which the voice is input,
A notification device for notifying an output from the output sound generation unit,
Connected to a communication network, the vehicle state information acquisition unit, the input voice recognition unit, the voice recognition data analysis unit, the operation execution determination unit, and a voice operation control device provided with the operation command generation unit,
Disposed on the vehicle, the voice input device, the notification device, and a voice operation input / output device provided with the device operation unit,
A communication interface device that is provided in each of the voice operation control device and the voice operation input / output device and connects the voice operation control device and the voice operation input / output device to be communicable via the communication network; Have.

  According to the configuration of (8), the functions of the functional unit of the vehicle voice operation device are distributed to the voice operation control device provided on the communication network side and the voice operation input / output device provided on the vehicle side. Specifically, a voice input device (such as a microphone) and an alarm device (such as a speaker) serving as an interface with a speaker such as a driver, and a device operation unit that executes an operation command for voice operation of an in-vehicle operation device include: The other functional units are arranged on the vehicle side, and the other functional units are arranged on the communication network side. The function unit provided in the voice operation control device is a function unit that can be shared by a plurality of vehicles, and the function unit shared by the plurality of vehicles is consolidated on the communication network side, thereby reducing vehicle cost. be able to. Further, by arranging the input voice recognition unit in the voice operation control device of the communication network, it is possible to collect many expression methods uttered from many vehicles in one place and collect the voice operation control device. It is possible to process various expression methods by referring to these data. As a result, it is possible to increase the reliability of the interactive exchange with the speaker and increase the reliability of the vehicular voice operation device.

(9) In some embodiments, in the configuration of the above (8),
The voice input device and the notification device are configured by a mobile communication device capable of performing data communication including a mobile phone, a smartphone, a tablet terminal, a notebook computer, an in-vehicle communication device, and a car navigation system.
According to the configuration (9), for example, the voice input device and the notification device are configured by an existing mobile communication device (mobile communication device) having a communication function using a mobile communication network, thereby reducing costs. Can be planned.

  According to at least one embodiment of the present invention, there is provided a vehicle voice operation device that controls execution of a voice operation in light of a state where a vehicle is placed.

It is a figure showing roughly composition of a vehicle provided with a voice operation device for vehicles concerning one embodiment of the present invention. It is a control block diagram of the voice operation device for vehicles concerning one embodiment of the present invention. FIG. 1 is a control block diagram of a vehicle voice operation device according to an embodiment of the present invention. The functions of the control block of the vehicle voice operation device are distributed between a vehicle and a cloud. It is a figure showing the control flow of the voice operation device for vehicles concerning one embodiment of the present invention. It is a figure showing the control flow of the voice operation device for vehicles concerning one embodiment of the present invention, and processes the response to a question message. It is a figure showing roughly composition of a vehicle provided with a voice operation device for vehicles concerning one embodiment of the present invention, and a voice operation control device provided in a communication network side and voice operation input / output including a mobile communication device The device is provided on the vehicle side.

  Hereinafter, some embodiments of the present invention will be described with reference to the accompanying drawings. However, the expression “comprising”, “comprising”, “including”, “including”, or “having” one component is not an exclusive expression excluding the existence of another component.

  FIG. 1 is a diagram schematically showing a configuration of a vehicle 6 including a vehicle voice operation device 1 according to one embodiment of the present invention. 2 and 3 are control block diagrams of the vehicle voice operation device according to one embodiment of the present invention. As shown in FIG. 1, the vehicle 6 has a vehicle voice operation device 1 and various sensors 3 mounted thereon, and is connected to each other via an on-vehicle network 61 (CAN, LIN, etc.), and thereby various types of sensors are provided. The detection result of the sensor 3 is input to the vehicle voice operation device 1. Further, each of the various sensors 3 is used to acquire vehicle state information I described later. For example, various sensors 3 include a rain sensor 31 that detects rainfall according to the state of raindrops on the window of the vehicle, an illuminance sensor 32 that detects illuminance outside the vehicle, an outside air temperature sensor 33 that detects an outside air temperature around the vehicle 6, A GPS sensor 34 for detecting the current position of the vehicle 6, a vehicle speed sensor 36, a camera 37 for detecting a situation in front of other vehicles and obstacles in front of the own vehicle 6, and a radar receiving an oscillating radio wave from a GPS satellite. Means 38 and the like may be included.

  The vehicle voice operation device 1 is a device that controls a voice operation of the vehicle-mounted operation device 4 mounted on the vehicle 6. Here, the in-vehicle operation device 4 is a device (equipment) that can be operated by a passenger, such as a headlight, a wiper, a blinker, an air conditioner, a power window, and a lock of a vehicle door. Then, the vehicle voice operation device 1 allows the occupant (speaker) of the vehicle 6 such as a driver to directly operate the operation device 42 such as a switch for operating the operation target device 4t, instead of the utterance of the occupant. To control the operation state of the operation target device 4t (ON state, OFF state, large, medium, small, stepwise operation levels such as high speed, medium speed, low speed, etc.). Therefore, as shown in FIGS. 2 and 3, the vehicle voice operation device 1 includes a vehicle state information acquisition unit 11, an input voice recognition unit 12, a voice recognition data analysis unit 13, and an operation execution determination unit 14. , An operation command generation unit 15, a device operation unit 16, and an output voice generation unit 17. Each functional unit will be described in order with reference to FIGS.

  The vehicle state information acquisition unit 11 acquires vehicle state information I indicating the state of the vehicle 6, as shown in FIGS. The vehicle state information I includes at least one of the operation state information Id, the vehicle side information Ii, and the external information Ie of the in-vehicle operation device 4 described below. The operation state information Id is information indicating whether the in-vehicle operation device 4 is operating or information indicating the operation level. The vehicle-side information Ii is the driving state of the vehicle 6 and information on the surroundings of the vehicle 6, and is information based on the detection result of the sensor 3 described above. Specifically, the driving state of the vehicle 6 is, for example, a vehicle speed. The peripheral information of the vehicle 6 includes the presence or absence of rain at the current position of the vehicle, the traveling environment such as night driving or running in a tunnel, the outside air temperature, the presence or absence of other vehicles and obstacles in front of the vehicle, and the like. In addition, the external information Ie includes the operation status information Id and the vehicle 6 information like the vehicle side information Ii, such as weather forecast, position information, traffic congestion information, road traffic information such as traffic regulation information based on accidents, construction, disasters, and weather conditions. Information that cannot be obtained inside the The above-mentioned position information is external information Ie because it is obtained by receiving a signal from a GPS satellite, receiving position information and a radio wave of a base station, and receiving other external information. The map information may use a car navigation system of the vehicle 6 or use a mobile communication device 8 such as a smartphone, and may be classified into at least one of the external information Ie and the vehicle-side information Ii.

  In the embodiment illustrated in FIGS. 2 and 3, the vehicle state information acquisition unit 11 acquires all of the above-described operation state information Id, vehicle-side information Ii, and external information Ie as vehicle state information I. More specifically, the operation state information Id is obtained by receiving the operation states of various on-board operation devices 4 from the device operation unit 16 described later. As described above, the vehicle-side information Ii is obtained by receiving detection values of various sensors 3 via the in-vehicle network 61. The external information Ie is acquired via the communication device 11v when the vehicle 6 includes a communication device 11v for communicating with the outside of the vehicle 6, which will be described later. The vehicle state information acquisition unit 11 stores the acquired vehicle state information I by including the database 11m.

  The input voice recognition unit 12 converts the input voice V into voice recognition data D. As shown in FIGS. 2 and 3, the input voice recognition unit 12 is connected to a voice input device 12v including a microphone for capturing voice V from a speaker. Then, the voice V from the speaker is input to the input voice recognition unit 12 via the voice input device 12v. In the embodiment shown in FIG. 2, the voice input to the input voice recognition unit 12 has a voice waveform (analog value), and the input voice recognition unit 12 compares the voice with an acoustic model or a language model in the voice recognition technology. Thus, the voice recognition data D is generated. In some other embodiments (such as FIG. 3 described later), data obtained by digitizing a voice waveform by a sampling technique may be input to the input voice recognition unit 12. That is, the voice V input to the input voice recognition unit 12 may be a digitized voice V. In this case, the digitized voice V may be inversely converted into a voice waveform, and a voice recognition technique may be applied to the voice waveform.

  The voice recognition data analysis unit 13 determines, based on the voice recognition data D, the in-vehicle operation device 4 to be the operation target device 4t and the operation content 4c for the operation target device 4t. As shown in FIGS. 2 and 3, the voice recognition data analysis unit 13 is connected to the input voice recognition unit 12 and receives the voice recognition data D output from the input voice recognition unit 12. Then, by analyzing the received voice recognition data, the content of the voice operation of the in-vehicle operation device intended by the speaker (operation content 4c for the operation target device 4t) is estimated, and the operation target device 4t and the operation content 4c are determined. I do. For example, when the utterance of the driver (speaker) is voice-recognized and voice recognition data D of "turn on the wiper" is obtained, the operation target device 4t is a wiper, and the operation content 4c is turned on ( Switch to the ON state). In this example, keywords indicating the operation target device 4t and the operation content 4c are included in the voice recognition data D, and the operation target device 4t and the operation content 4c are directly determined (guessed) from the voice recognition data D. This is an example when it is possible.

  On the other hand, for example, when the voice of the driver (speaker) is voice-recognized and voice recognition data D indicating that "rain has stopped" is obtained, the operation target device 4t is a wiper, and the operation content 4c is OFF ( Switch to the off state). That is, in this example, the keywords indicating the operation target device 4t and the operation content 4c are not included in the voice recognition data D. Therefore, in advance, keywords indicating conditions such as “rain has stopped”, “rain has fallen”, “dark”, “bright”, “cold”, “hot” and in-vehicle operation that the keywords are related to Associative data associated with the device 4 is prepared. Then, by referring to the associative data, the voice recognition data analysis unit 13 determines that the speaker intends to turn off the wiper from the keywords “rain” and “stopped” included in the voice recognition data D. presume. Based on the content estimated in this way, the voice recognition data analysis unit 13 determines (guess) the operation target device 4t and the operation content 4c. That is, in this case, the operation target device 4t is a wiper, and the operation content 4c is turned off (switched to the off state).

  Further, as described later, in some embodiments, the vehicle voice operation device 1 may notify the question message Q. Then, when the voice V responding to the question message Q is input, the voice recognition data analysis unit 13 outputs, for example, voice recognition indicating a positive response such as “high” and a negative response such as “no”. Data D is input. At this time, the voice recognition data analysis unit 13 prepares the expected value (operation target device 4t, operation content 4c, and the like) in the case of an affirmative response or a negative response by asking the voice recognition data D based on the voice recognition data D. Then, the in-vehicle operation device 4 serving as the operation target device 4t and the operation content 4c for the operation target device 4t are determined.

  In some embodiments, the voice recognition data analysis unit 13 may refer to the vehicle state information I from the vehicle state information acquisition unit 11 when analyzing the voice recognition data D. Thereby, the voice recognition data analysis unit 13 recognizes the operation target device 4t from the voice recognition data D, and determines the operation content 4c in consideration of the vehicle state information I associated with the operation target device 4t. Then, the operation target device 4t and the operation content 4c are determined based on the voice recognition data D. For example, the rain sensor 31 is associated with the wiper in advance by the association data, and by referring to the association data, the rain condition is determined based on the rain sensor 31 from the voice recognition data D of “wiper”. Then, it is possible to perform the same voice operation as “turn on the wiper”.

  The operation execution determination unit 14 determines the validity of executing the operation content 4c on the operation target device 4t based on the vehicle state information I. As shown in FIGS. 2 and 3, the operation execution determination unit 14 is connected to the voice recognition data analysis unit 13 and receives the operation target device 4t and the operation content 4c from the voice recognition data analysis unit 13. The operation execution determination unit 14 is also connected to the vehicle state information acquisition unit 11, receives the vehicle state information I acquired by the vehicle state information acquisition unit 11, and recognizes the state where the vehicle is placed. Based on the vehicle state information I, the operation execution determination unit 14 does not execute the voice operation according to the utterance content as it is, but determines whether or not it is appropriate to execute the voice operation. And a configuration for controlling the execution of the voice operation. That is, as described later, when the operation execution determination unit 14 determines that there is validity, the operation execution determination unit 14 executes the operation content 4c on the operation target device 4t (performs a voice operation), and conversely, the validity is determined. When it is determined that there is no operation, the operation content 4c is not executed on the operation target device 4t (no voice operation is executed). Then, as described later, the operation execution determination unit 14 generates notification data N including a response message R indicating that there is no validity, or generates an operation command C, based on the validity determination result. .

  The operation command generation unit 15 generates an operation command C for executing the operation content 4c on the operation target device 4t according to the instruction of the operation execution determination unit 14. As shown in FIGS. 2 and 3, the operation command generation unit 15 is connected to the operation execution determination unit 14. Then, the operation command generation unit 15 receives the instruction from the operation execution determination unit 14 and executes a voice operation on the operation target device 4t. Specifically, when the operation execution determination unit 14 receives the operation target device 4t and the operation content 4c from the operation instruction generation unit 15, the operation execution determination unit 14 may generate the operation instruction C in response to the reception.

  The device operation unit 16 executes the operation command C. As shown in FIGS. 2 and 3, the device operation unit 16 is connected to the operation command generation unit 15. Then, upon receiving the operation command C, the device operation unit 16 operates the operation target device 4t connected via the in-vehicle network 61. Note that an ECU (electronic control unit) (not shown) that controls the operation target device 4t may be present. In this case, the device operation unit 16 issues an operation command C via these ECUs (not shown). Execute.

  In the embodiment illustrated in FIGS. 2 and 3, the device operation unit 16 transmits the operation state of the operation target device 4t to the vehicle state information acquisition unit 11. Specifically, a change in the operation state of one or more operation target devices 4t is monitored, and the operation state of the operation target device 4t in which the change in the operation state is detected is voluntarily transmitted to the vehicle state information acquisition unit 11. doing. Then, every time the vehicle state information acquisition unit 11 receives the operation state, it changes the management information of the database 11m, and maintains the operation state information Id at the latest state. In some other embodiments, the device operation unit 16 may acquire the operation state in response to a request from the vehicle state information acquisition unit 11.

  The output sound generation unit 17 converts the notification data N received from the operation execution determination unit 14 into a sound V. As shown in FIGS. 2 and 3, the output sound generation unit 17 is connected to the operation execution determination unit 14 and a notification device 17v such as a speaker. Then, the output sound generation unit 17 converts the notification data N received from the operation execution determination unit 14 into a sound V, and outputs a control result for the input sound to the vehicle 6 via the notification device 17v (such as a speaker) (in the vehicle 6). (Passenger).

  In the above configuration, the operation execution determination unit 14 instructs the operation instruction generation unit 15 to generate the operation instruction C when the operation execution determination unit 14 determines that the operation instruction C is valid. If it is determined, the notification data N for outputting a response message R indicating that the response is invalid is output to the output sound generation unit 17 (see FIGS. 4 and 5). That is, when the operation command generation unit 15 is instructed to generate the operation command C, the voice operation on the operation target device 4t is performed, and conversely, the above-mentioned invalidity is indicated. When the notification data N including the response message R is output to the output voice generation unit 17, the voice operation on the operation target device 4t is not executed or the execution is suspended.

Next, a control flow of voice operation in the vehicle voice operation device 1 will be described with reference to FIGS. 4 and 5 are diagrams showing a control flow of the vehicle voice operation device 1 according to one embodiment of the present invention.
In step S1 of FIGS. 4 and 5, when a voice of a speaker such as a driver of the vehicle 6 is input, the vehicle voice operation device 1 analyzes the input voice recognition data D. In step S2 of FIGS. 4 and 5, the vehicle voice operation device 1 determines the operation target device 4t and the operation content 4c based on the voice recognition data D. In step S3 of FIGS. 4 and 5, the vehicle voice operation device 1 determines the validity of executing the operation content 4c on the operation target device 4t based on the vehicle state information I. If it is determined in step S4 of FIGS. 4 and 5 that the vehicle voice operation device 1 has the validity, the operation of the operation target device 4t is performed in step S5 of FIGS. Execute content 4c.

  Thereafter, in the embodiment of FIGS. 4 and 5, the execution result of the voice operation in step S5 is notified by voice, and then the control flow ends. For example, when the voice recognition data D that has been voice-recognized has a content of "attach a wiper", a voice V such as "attached a wiper" is notified. It should be noted that the notified voice V is a voice of the notification data N generated by the operation execution determination unit 14 described above, and the operation execution determination unit 14 outputs the voice V “Wiper attached”. Is generated, and is output to the output sound generation unit 17.

  On the other hand, a case will be described in which the voice operating device 1 for a vehicle determines that the above-mentioned validity is not satisfied in step S4 of FIGS. In the embodiment shown in FIG. 4, in step S7a, the invalidity is notified by the sound V, and then the control flow ends. On the other hand, in the embodiment shown in FIG. 5, at step S7b, a response message R and a question message Q (described later) indicating invalidity are notified by voice V. Thereafter, the control flow for the current input voice V ends, but the vehicle voice operation device 1 monitors whether the next input voice V is a response to the question message Q. Specifically, in the embodiment shown in FIG. 5, step S8 exists between step S1 and step S2 in FIG. 5 described above. In step S8, when the voice operation device 1 for a vehicle determines that the response is to the question message Q as a result of analyzing the input voice recognition data D (step S1), the content of the response is determined in step S9. Control according to the following. In the embodiment shown in FIG. 5, the result of the control according to the content of the response is reported.

  Step S9 in FIG. 5 will be described using the above example of “rain stopped”. For the first voice recognition data D indicating “rain stopped”, “Would you like to stop the wiper?” Is sent. At this time, the vehicular voice operation device 1 is to switch the operation target device 4t to a wiper and to switch the operation content 4c to an off state in response to a positive response, and conversely, to respond to a negative response. For example, the operation target device 4t is a wiper, and the operation content 4c is prepared with an expected value such as switching to the ON state. If the voice input after the question message Q is notified is determined to be a response voice such as "high" or "no", the process proceeds to step S9 in FIG. Then, it is determined whether the response voice is a positive response or a negative response, and the expected value is executed according to the result. If it is not determined that the response is a response in step S9 of FIG. 5, it is considered that the question message Q has been ignored, and therefore the execution of the expected value for the positive response, the execution of the expected value for the negative response, or the voice Any of not performing the operation may be performed. Then, the input voice V may be subjected to step S2 in FIG. 5 as a new voice operation.

  According to the above configuration, the vehicular voice operation device 1 captures the voice (utterance) of the driver of the vehicle 6 and the like, and the content of the voice operation intended by the speaker (operation target device and operation content) by voice recognition. Is determined (guessed). Then, the vehicle voice operation device 1 performs a voice operation (operation command C) for reflecting the voice operation on the vehicle in light of the current vehicle state (vehicle state information I) where the vehicle 6 is placed. ) Is determined to be appropriate. Then, when it is determined that the execution of the voice operation is not appropriate, the driver is notified of the fact without performing the voice operation. By controlling the execution of the voice operation in accordance with the state of the vehicle in this way, for example, it is possible to prevent a dangerous state from occurring, such as performing a voice operation such as turning off a headlight while traveling in a tunnel, It can support safe driving.

  In addition, as described with reference to FIG. 5 described above, in some embodiments, the notification data N includes a question message Q regarding the execution of the operation content 4c together with a response message R indicating that the notification data is not valid, and performs voice recognition. When receiving a response to the question message Q via the input voice recognition unit 12, the data analysis unit 13 issues an instruction (an instruction to generate the operation instruction C) to the operation instruction generation unit 15 according to the content of the response. Execute or do not execute the operation content 4c on the operation target device 4t. According to the embodiment shown in FIG. 5 described above, when the above-mentioned invalidity is not satisfied, the question message Q is notified, and the execution of the voice operation is controlled according to the response of the speaker, so that the intention related to the voice operation of the speaker is performed. Is confirmed. As a result, the accuracy of the interactive exchange with the speaker can be increased, and the reliability of the vehicle voice operation device 1 can be increased.

  Further, in some embodiments, as described above, the vehicle 6 includes the sensor 3 for detecting at least one of the driving state of the vehicle 6 and information on the surroundings of the vehicle 6 as the vehicle-side information Ii. Information I includes vehicle-side information Ii. Then, the operation execution determination unit 14 determines the validity of executing the operation content 4c on the operation target device 4t based on the vehicle state information I including the vehicle side information Ii. Specifically, when it is determined that execution of the operation command C for the operation target device 4t does not conform to the vehicle side information Ii associated with the operation target device 4t in light of the function of the operation target device 4t, Is determined to be invalid, and when it is determined to be compatible with the vehicle-side information Ii associated with the operation target device 4t, it is determined that the above is valid. That is, the vehicle voice operation device 1 has the above-described association data for indicating what the sensor 3 that detects information related to each of the vehicle-mounted operation devices 4 is. Then, the operation execution determination unit 14 refers to the association data, and based on a comparison between the vehicle-side information Ii related to the operation target device 4t determined from the input voice V and the operation content 4c, Determine the validity of performing the voice operation.

  More specifically, it is assumed that the headlight and the illuminance sensor 32 are associated with each other in the above association data. At this time, for example, when the input voice V has the content “turn off the headlights”, the operation target device 4t is the headlight, and the operation details 4c are switched to the off state. Here, if the function of the headlight is to illuminate the front in the traveling direction of the vehicle 6 in the dark, it is dangerous to turn off the headlight at night or when the vehicle 6 is traveling in a tunnel. . In order to prevent the vehicle 6 from falling into such a dangerous state, the vehicular voice operation device 1 includes the above-described association data. Then, the operation execution determination unit 14 refers to the vehicle-side information Ii regarding the illuminance sensor 32 related to the function of the headlight by referring to the association data for the content of the voice operation.

  If the detection result of the illuminance sensor 32 is smaller than a predetermined illuminance value set as a state in which the headlights need to be turned on, the above-mentioned voice operation is replaced with a voice operation on the headlights (switching to an off state). Does not match and is determined to be invalid. If it is determined that the response message R is not appropriate, for example, a response message R such as “It is dangerous if the headlight is turned off now, so it automatically turns off when it becomes bright” may be reported as invalid. Alternatively, it may be notified that the response message R such as “It is dangerous if the headlights are now turned off, so that they are automatically turned off after exiting the tunnel.” Is invalid. This tunnel can be determined based on position information (external information) and map information. Conversely, when the detection result of the illuminance sensor 32 is larger than the predetermined illuminance value, the voice operation device 1 for a vehicle is suitable for the voice operation of the headlight (switching to the off state). And determine that it is appropriate. Thereby, a voice operation is performed.

  Further, it is assumed that the wiper and the rain sensor 31 are associated with each other in the above association data. At this time, for example, if the input voice V has the content "rain stopped", the operation target device 4t is a wiper, and the operation content 4c is switched to the off state. Here, the function of the wiper is to wipe off the raindrops (moisture) on the window glass (for example, the windshield) of the vehicle 6. If there is no raindrop on the window glass, it is not necessary to operate the wiper. poor. When it is determined that such a voice operation is unnecessary, the vehicular voice operation device 1 refers to the association data for the content of the voice operation, and the vehicle related to the rain sensor 31 related to the wiper function. Reference is made to the side information Ii. Then, when the detection result of the rain sensor 31 is larger than a predetermined rainfall amount set as a state necessary for stopping the wiper, the above-mentioned voice operation conforms to the operation content 4c (switch to the off state) for the wiper. No, it is judged to be invalid. When it is determined that the response message R is not appropriate, for example, a response message R such as “Rainfall has been detected. Conversely, when the detection result (for example, no rainfall) of the rain sensor 31 is smaller than the predetermined rainfall amount, the voice operation device 1 for a vehicle conforms to the operation content 4c for the wiper. Is determined to be valid. Thereby, a voice operation is performed.

  Note that the sensor 3 may be configured to be switched between an on state where the detection by the sensor 3 is performed and an off state where the detection is not performed by the sensor 3. In this case, the notification content may be determined according to the ON / OFF state of the sensor 3. For example, in the case where the input voice V has the content "rain stopped", if the rain sensor 31 is off, it is not possible to detect rainfall. For this reason, a response message R and a question message Q indicating invalidity may be reported, such as "If the rain has stopped, should the wiper be stopped?"

  According to the above configuration, the vehicle state information I includes at least the vehicle side information Ii, and the content of the voice operation of the in-vehicle operation device 4 is related to the operation target device 4t obtained by the sensor 3 installed in the vehicle 6. The validity of the vehicle 6 is determined by comparison with the state of the vehicle 6. As a result, for example, it is possible to prevent the operation command C such as turning off the headlight while the vehicle 6 is traveling in the tunnel from being executed, and to assist safe driving.

  In some embodiments, as shown in FIG. 1, the vehicle 6 includes a communication device 11v for communicating with the outside of the vehicle 6, and the vehicle state information I is obtained via the communication device 11v. The external information Ie is included. Then, the operation execution determination unit 14 determines the validity of executing the operation content 4c on the operation target device 4t based on the vehicle state information I including the external information Ie. Specifically, based on the external information Ie, an operation command C that is the reverse of the operation command C for the operation target device 4t is executed on the operation target device 4t within a predetermined time after the execution of the operation instruction C for the operation target device 4t. If the prediction of the fact is established, it is determined that there is no validity described above, and if the prediction is not satisfied, it is determined that there is the validity described above. The communication device 11v is a device that enables communication between the vehicle 6 and the information providing server 92 that provides the external information Ie connected to the communication network 9 (wired network) wirelessly, and the vehicle state information. The acquisition unit 11 acquires the external information Ie via the communication device 11v. The vehicle state information I is configured to include at least the external information Ie, and the vehicle voice operation device 1 determines the validity of the execution of the voice operation determined from the input voice V by the external information Ie. Is determined based on

  More specifically, for example, when the input voice V has a content of “attach a wiper”, the operation target device 4t is a wiper, and the operation content 4c is switched to an on state. become. If the voice operation is performed immediately before entering the tunnel, the vehicle 6 enters the tunnel after executing the voice operation, and the content of the immediately executed voice operation is reversed. It is expected that voice operations such as "te" will be performed. When it is expected that the reverse operation on the operation target device 4t will be performed continuously, it is convenient for the driver and the like not to execute the first voice operation. There are cases.

  Therefore, the vehicle voice operation device 1 refers to the position information and the map information (such as a car navigation system) of the vehicle 6 to obtain information ahead of the vehicle (in this example, the presence or absence of a tunnel) for the content of the above voice operation. get. Then, the time when the vehicle 6 arrives at the acquired front tunnel or the like is estimated from the detection value of the vehicle speed sensor 36 or the like, and if this time is within a predetermined time set as a case where the above-described possibility occurs. In this case, the vehicle voice operation device 1 determines that the execution of the voice operation is not appropriate. If it is determined that the response message R is not valid, for example, a response message R such as “There is a tunnel ahead. Should you exit the tunnel?” May be reported as invalid. Alternatively, additional information that the speaker is expected to know may be added to the response message R. Specifically, the content such as "The weather at the destination at the estimated arrival time is fine." May be notified in addition to the response message R described above. Conversely, if the estimated time is not within the predetermined time, the voice operation device 1 for a vehicle determines that the voice operation of “turn on the wiper” is appropriate. Thereby, a voice operation is performed.

  According to the above configuration, the peripheral information viewed from the position information of the vehicle 6 and the external information Ie such as traffic congestion information and construction information around the route to the destination are acquired using the communication device 11v. Further, the validity of the content of the voice operation of the in-vehicle operation device 4 is determined by being compared with the external information Ie. For example, when there is a voice operation indicating that the wiper is activated, if it is predicted from the external information Ie that the vehicle will enter the tunnel immediately after the execution of the voice operation, the wiper is activated immediately before the tunnel. However, it is expected that a voice operation to stop the wiper will be performed by immediately entering the tunnel. As described above, even if the first operation command C is executed according to the intention of the speaker, it is predicted that the opposite operation command C is immediately executed by the second voice operation. The operation device 1 is configured not to immediately execute the operation command C for the operation target device 4t. Thus, it is possible to prevent a situation in which the speaker repeats the voice operation, and to support a safe driving by forming a situation in which the driver can concentrate on driving.

  Further, in some embodiments, as described above, the vehicle state information I includes the operation state information Id of the vehicle-mounted operation device 4. Then, the operation execution determination unit 14 determines the validity of executing the operation content 4c on the operation target device 4t based on the vehicle state information I including the operation state information Id of the operation target device 4t. Specifically, when it is determined that the operation state information Id of the operation target device 4t and the operation content 4c match, it is determined that there is no validity, and the operation state information Id of the operation target device 4t is determined. If it is determined that the operation does not match the operation content 4c, it is determined that the above-described validity exists. That is, the vehicle state information I is configured to include at least the operation state information Id, and the vehicle voice operation device 1 determines whether the execution of the voice operation determined from the input voice V is appropriate for the operation target. The determination is made based on a comparison between the operation state information Id of the device 4t and the operation content 4c.

  More specifically, for example, when the input voice V has a content of “attach a wiper”, the operation target device 4t is a wiper, and the operation content 4c is switched to an on state. become. Upon receiving the voice operation, the vehicle voice operation device 1 checks the operation state information Id of the wiper to determine whether the wiper is operating or stopped. Then, when the operation state information Id of the wiper is in operation, the vehicle voice operation device 1 determines that the operation content 4c matches the operation content 4c of switching to the on state, and the voice operation is valid. It is determined that there is not. If it is determined that the response message R is not valid, for example, a response message R such as “The wiper is already moving” may be reported as invalid. Conversely, when the operation state information Id of the wiper is stopped, it is determined that the operation content 4c does not match the operation content of switching to the on state, and that the voice operation is appropriate. Thereby, a voice operation is performed.

  According to the above configuration, the validity of the voice operation of the in-vehicle operation device 4 is determined by being compared with the operation state of the in-vehicle operation device 4. Thus, for example, in a state in which the wiper is operating, it is possible to prevent an unnecessary load such as executing an operation command for operating the wiper from being applied to the device operation unit 16, and the response of the vehicle voice operation device 1 is reduced. Can be enhanced.

  Further, in some embodiments, as in the embodiments shown in FIGS. 1 to 5, the operation content 4c is executed on the operation target device 4t in addition to the operation status information Id in the vehicle status information I. If other information for determining the validity of the operation is included, the above-described validity is determined based on the operation state information Id of the operation target device 4t. Only the validity determination based on other information is further performed. That is, in the embodiment shown in FIGS. 1 to 5, the vehicle voice operation device 1 is activated before the validity of the voice operation based on the vehicle side information Ii or the external information Ie as other information. It is configured to determine the validity of the voice operation based on the state information Id. If the validity of the voice operation is determined to be invalid based on the operation state information Id, the validity of the voice operation is determined based on the vehicle-side information Ii or the external information Ie. It is determined that there is no validity of the voice operation. If it is determined that the voice operation is valid based on the operation state information Id, the validity of the voice operation based on the vehicle-side information Ii and the external information Ie is performed.

  In the embodiment shown in FIGS. 1 to 5, the vehicle state information I includes vehicle-side information Ii, external information Ie, and operation state information Id. In this case, for example, information similar to the vehicle-side information Ii and the external information Ie, such as weather information such as the presence or absence of rainfall at the current position of the vehicle 6 and road traffic information (the situation in front of the vehicle 6), is obtained. In some cases. For example, when rainfall is detected by the rain sensor 31 but the weather forecast from the information providing server 92 indicates opposite contents, such as when the weather is fine, the vehicle-side information Ii having higher reliability is used as the information. By setting a priority such as a priority over the external information Ie, it is possible to more appropriately determine the validity of the voice operation.

  According to the above configuration, when it is determined that the voice operation based on the operation state information Id of the operation target device 4t is appropriate, the voice operation based on other information such as the vehicle side information Ii and the external information Ie is performed. Is determined. Conversely, if it is determined that the voice operation is not valid based on the operation state information Id, the validity of the voice operation is determined based on other information (the vehicle-side information Ii and the external information Ie). Therefore, it is determined that the voice operation is not valid. The determination of the validity of the voice operation based on the operation state information Id of the operation target device 4t only needs to confirm the operation state of the operation target device 4t, and the determination criteria of the vehicle side information Ii, the accuracy of the external information Ie, and the like. Since it does not depend on the passenger's personal feeling or situation, it can be performed more accurately than other information. For this reason, the accuracy of the interactive exchange with the speaker can be increased, the response of the vehicle voice operation device 1 can be improved, and the reliability of the vehicle voice operation device 1 can be improved.

  Further, in some embodiments, the in-vehicle operating device 4 is configured to be able to set the operation state stepwise, and the operation state information Id includes setting step information Il indicating a setting step of the operation state. Contains. When the operation content 4c for the operation target device 4t is determined to be an operation for relatively changing the setting state of the operation state, the operation execution determination unit 14 is obtained by the vehicle state information obtaining unit 11. The validity of executing the operation content 4c on the operation target device 4t is determined based on the vehicle state information I including the setting stage information Il. For example, in the on-board operation device 4 such as the speed of the wiper, the irradiation angle of the headlight, the temperature setting of the air conditioner, the vacancy of the window, etc., in addition to the ON state and the OFF state, the operating state is set stepwise in the ON state. There are things you can do. Then, there is a case where a passenger, such as a driver, wants to operate the operating state of the vehicle-mounted operation device 4 faster and lower based on the current set value. Taking the speed of the wiper as an example, some wipers have an operation level of multiple stages (for example, five stages), and the operation level can be changed from a low speed to a high speed to each state of 1 to 5. . In addition, the driver or the like may not only perform the voice operation by indicating the operation level, but also desire an operation such as earlier or later depending on the current rainfall condition. When relative operation is enabled, convenience is increased.

  Therefore, in the embodiment shown in FIGS. 1 to 5, the voice operation device for a vehicle 1 is configured so that the operation of the operation level of the wiper can be relatively changed from the current operation level. For example, the operation level is defined to be two steps earlier for the operation command C of “early”, and the operation level is decreased by two steps for the operation command C of “slow”. In addition, in response to the operation command C of “slightly” that works to weaken the operation of “early” such as “slightly faster”, the operation level is increased by one step from the current level, and “faster” such as “faster” In response to the operation command "More" that works to enhance the operation, the operation level is set to be three stages earlier than the current level. Similarly, in response to the operation command C of "slightly", which acts to weaken the operation of "slowly", such as "slightly late", the operation level is delayed by one step from the current state, and "slower", such as "slower". In response to the operation command C of "more" which works to strengthen the operation of "", the operation level is delayed by three stages from the current state.

  In this case, the vehicle voice operation device 1 determines the validity of the voice operation regarding the relative change of the operation level of the wiper serving as the operation target device 4t by referring to the vehicle side information Ii and the external information Ie. . For example, when it is determined that the operation content 4c is not suitable in light of the rainfall amount by the rain sensor 31, it is determined that there is no validity. Specifically, the operation level of the wiper with respect to the rainfall amount is defined in the above-described association data and the like, and a desired operation level can be determined by comparing the rainfall amount with a threshold value according to the operation level. I have. Then, by referring to the association data and the like, when it is desirable to operate the wiper at the maximum speed with extremely strong rain legs, for example, when the current operation level of the wiper is 2, If the voice operation is performed one step earlier, such as "slightly advance", the operation level of the wiper does not become the highest even if the voice operation is performed. In such a case, the vehicle voice operation device 1 determines that the voice operation is not appropriate. If it is determined that the voice operation is not appropriate, the response message R indicating that an alternative such as “the wiper operation has been set to the highest level” may be notified as invalid. Further, since the vehicle voice operation device 1 does not follow the content of the voice operation performed by the speaker, after setting the highest level, a question such as "The wiper operation has been set to the highest level. Is there any problem?" You may. In this case, if the response to the question is affirmative or no response, the content of the operation may be maintained, and if the response is negative, the content may be changed again according to the content of the utterer's operation (that is, the wiper is made slightly faster).

  Further, for example, in the temperature setting of the air conditioner, the set temperatures are increased by 1 ° C., 2 ° C., 3 ° C., and “low”, For voice expressions such as "Slightly lower" and "Much lower", if the voice operation is possible to lower the set temperature by 1 ° C, 2 ° C, and 3 ° C, respectively, the validity is the same as in the example of the wiper described above. Is determined. Specifically, the set temperature of the air conditioner with respect to the outside air temperature is defined in the above-mentioned association data and the like. By comparing the outside air temperature with a threshold value according to the set temperature, a desired set temperature (appropriate set temperature) is obtained. You can decide. When a voice operation is performed to lower (increase) the set temperature of the air conditioner, a temperature difference between the set temperature obtained by comparing the associated data with the outside air temperature and the temperature to be set by the voice operation is predetermined. If the temperature difference is larger than the temperature difference, it is determined that the voice operation is not appropriate. If it is determined that it is not appropriate, for example, it is notified that the operation content 4c of the alternative has been executed, such as “The temperature difference from the outside air temperature is X ° C., so lower it a little.” A question such as “The temperature difference from the outside air temperature is X ° C., should the temperature be lowered by 3 ° C. (increase)?” May be reported as invalid.

  Similarly, when a preceding vehicle is traveling in a predetermined range ahead, when a voice operation such as switching a headlight to a high beam is performed, it is determined that the voice operation is not appropriate, and execution of an alternative solution or questioning is performed. May be reported as invalid. In addition, in the operation of the window of the power window, for example, when the air is dirty and it is not appropriate to open the window, the execution of the alternative or the notification that the question is not appropriate may be made.

  Note that, when the operation content 4c for changing the operation level of the operation target device 4t such as a wiper is executed by the vehicular voice operation device 1, if the operation level cannot be changed as defined, the validity of the voice operation is determined. If it is determined that there is no voice operation and the operation level can be changed according to the operation content 4c, it may be determined that the voice operation is appropriate. For example, assume that the wiper has five operating levels. In this case, when the current operation level is set to the third stage, when the operation content 4c of “faster” is executed, the operation level that the wiper does not have indicates the sixth stage. Therefore, it is determined that it is not appropriate. If it is determined that the response message R is not appropriate, for example, the response message R that notifies the execution of the alternative such as “the operation of the wiper has been set to the highest level” may be notified as invalid. Further, a question message such as before setting the highest level or whether the highest level may be set may be added to the response message R and notified.

  According to the above configuration, the voice operation such as changing the operation stage of the operation target device 4t, such as the strength of the wiper, is compared with the state of the vehicle 6 obtained by the sensor 3 installed in the vehicle 6. , Its validity is determined. For example, when it is determined that it is appropriate to maximize the operation level of the wiper due to heavy rain legs and the content of the voice operation does not conform to this determination, it is determined that the above-described validity is not obtained. Thereby, the in-vehicle operation device 4 can be guided to an appropriate operation state according to the traveling state of the vehicle 6, and safe driving can be supported.

  In some embodiments, as shown in FIG. 3, the vehicle voice operation device 1 includes a voice operation control device 72 connected to the communication network 9 and a voice operation input / output device disposed in the vehicle 6. And a communication interface device 18 provided in each of the voice operation control device 72 and the voice operation input / output device 62 for communicably connecting the voice operation control device 72 and the voice operation input / output device 62 via the communication network 9. (18c, 18m), a voice input device 12v for capturing an input voice V, and a notification device 17v for notifying an output from the output voice generation unit 17. The voice input / output device 62 on the vehicle 6 side includes a voice input device 12v, a notification device 17v, and a device operation unit 16. On the other hand, the voice operation control device 72 on the communication network 9 side includes a vehicle state information acquisition unit 11, an input voice recognition unit 12, a voice recognition data analysis unit 13, an operation execution determination unit 14, and an operation command generation unit 15. Can be The output voice generation unit 17 is provided in one of the voice operation control device 72 and the voice operation input / output device 62. Then, by the communication interface device 18, between the voice input device 12v and the input voice recognition unit 12, between the notification device 17v and the output voice generation unit 17, between the device operation unit 16 and the operation command generation unit 15, and The device operation unit 16 and the vehicle state information acquisition unit 11 are connected to each other. That is, the functions of the vehicle voice operation device 1 are distributed to the voice operation control device 72 and the voice operation input / output device 62, and the voice operation control device 72 and the voice operation input / output device are connected via the communication interface device 18. The communication with the output device 62 controls execution of the voice operation.

  In the embodiment shown in FIG. 3, the voice operation control device 72 is provided on the cloud 7. In other words, it is configured by using a server group and a storage group constituting the cloud 7, and the software (program) operates in the server group, so that the vehicle state information acquisition unit 11, the input voice recognition unit 12, The functions of the functional units of the voice recognition data analysis unit 13, the operation execution determination unit 14, the operation command generation unit 15, and the output voice generation unit 17 are realized. Further, the database 11m in which the vehicle state information I is stored is realized by a storage group constituting a cloud. Thus, the vehicle state information I managed for each vehicle 6 in the embodiment shown in FIG. 2 is collected and managed in the voice operation control device 72 in the embodiment shown in FIG.

  As shown in FIG. 3, the voice input device 12v arranged on the vehicle 6 side digitizes voice V (voice waveform) input via a microphone or the like by a sampling technique. , And converts the input voice V into voice data Vd. Thereafter, communication data is generated by, for example, converting the voice data Vd into an IP packet, and the communication data is transmitted to the communication network 9 via the communication interface device 18m on the vehicle 6 side. This communication data is routed through the communication network 9 and received by the communication interface device 18c of the voice operation control device 72, and then is passed through the voice conversion unit 12d for returning the voice data Vd to the voice V (voice waveform). Then, the voice V is input to the input voice recognition unit 12.

  The output voice generation unit 17 is provided in one of the voice operation control device 72 and the voice operation input / output device 62. In the embodiment illustrated in FIG. 3, the output voice generation unit 17 is provided in the voice operation control device 72 (on the communication network 9 side), and the voice V in which the notification data N is voiced by the output voice generation unit 17 is The audio V (audio waveform) is converted into audio data Vd via an audio conversion unit 12d for digitizing the audio V by a sampling technique. Then, for example, the voice data Vd is converted into an IP packet to generate communication data, and the communication data is transmitted to the communication network 9 via the communication interface device 18m on the vehicle 6 side. This communication data is routed through the communication network 9, is received by the communication interface device 18m of the voice operation input / output device 62, and is then input to the notification device 17v, so that the voice data Vd is converted into voice V. , A speaker or the like. In some other embodiments, the output voice generation unit 17 is provided in the voice operation input / output device 62 (on the vehicle 6 side), and the notification data N output from the operation execution determination unit 14 is directly transmitted to the IP packet. The communication data is converted into communication data such as data and transmitted from the communication interface device 18 m on the vehicle 6 to the communication network 9. Then, the communication data is input from the communication interface device 18m on the vehicle 6 side to the output sound generation unit 17, is converted into the sound V, and is notified from the notification device 17v.

  Further, between the device operation unit 16 on the vehicle 6 and the operation command generation unit 15 on the communication network 9 and between the device operation unit 16 on the vehicle 6 and the vehicle state information acquisition unit 11 on the communication network 9. Then, the operation command C and the operation state information Id are directly converted into communication data such as an IP packet, and are exchanged via the communication network 9 as described above.

  According to the above configuration, the functions of the functional unit of the vehicle voice operation device 1 are distributed to the voice operation control device 72 provided on the communication network 9 side and the voice operation input / output device 62 provided on the vehicle 6 side. Specifically, the voice input device 12v (such as a microphone) and the notification device 17v (such as a speaker) serving as an interface with a speaker such as a driver, and the operation command C for voice operation of the vehicle-mounted operation device 4 are executed. The device operation unit 16 is arranged on the vehicle 6 side, and the other functional units are arranged on the communication network 9 side. The functional unit provided in the voice operation control device 72 is a functional unit that can be shared by a plurality of vehicles, and the functional units shared by the plurality of vehicles are aggregated on the communication network 9 side, so that the Cost can be reduced. In addition, by arranging the input voice recognition unit 12 in the voice operation control device 72 of the communication network 9, it is possible to collect many expression methods uttered from many vehicles 6 in one place, The voice operation control device 72 can process various expression methods by referring to these data. As a result, the reliability of the interactive exchange with the speaker can be increased, and the reliability of the vehicle voice operation device 1 can be increased.

  In some embodiments, as shown in FIG. 6, the voice input device 12v and the notification device 17v may be a data communication device including a mobile phone, a smartphone, a tablet terminal, a notebook computer, an in-vehicle communication device, and a car navigation system. Is constituted by a mobile communication device 8 capable of performing the following. FIG. 6 is a diagram schematically illustrating a configuration of a vehicle including the vehicle voice operation device according to one embodiment of the present invention. As shown in FIG. 6, the voice operation control device 72 is provided on the communication network 9 side, and the voice operation input / output device 62 including the mobile communication device 8 is provided on the vehicle side. The mobile communication device 8 establishes a wireless connection with a wireless base station existing around the current position of the vehicle 6, and communicates with the voice operation control device 72 via the wireless connection. In the mobile communication device 8, dedicated software (application) is running, and communication between the dedicated software and the voice operation control device 72 as a communication destination is performed on the above wireless connection. . The external information Ie acquired by the mobile communication device 8 may be sent to the vehicle state information acquisition unit 11. In addition, the mobile communication device 8 is connected to the voice operation control device 72 by a short-range communication standard such as Bluetooth (registered trademark), wireless such as WiFi, or a wire such as a communication cable, so that the voice operation control device 72 is connected. It may communicate with the voice operation control device 72 via the communication interface device 18m on the side.

  According to the above configuration, for example, the cost can be reduced by configuring the voice input device and the notification device using an existing mobile communication device (mobile communication device) having a communication function using a mobile communication network. it can.

  The present invention is not limited to the above-described embodiment, and includes a form in which the above-described embodiment is modified and a form in which these forms are appropriately combined.

Reference Signs List 1 vehicle voice operation device 11 vehicle state information acquisition unit 11m database 11v communication device 12 input voice recognition unit 12v microphone 12d voice conversion unit 13 voice recognition data analysis unit 14 operation execution determination unit 15 operation command generation unit 16 device operation unit 17 output Voice generator 17v Speaker 18 Communication interface device 3 Sensor 31 Rain sensor 32 Illuminance sensor 33 Outside air temperature sensor 36 Vehicle speed sensor 37 Camera 38 Radar means 4 In-vehicle operation device 42 Operation device 4c Operation content 4t Operation target device 6 Vehicle 61 In-vehicle network 62 Voice Operation input / output device 7 Cloud 72 Voice operation control device 8 Mobile communication device 9 Communication network 92 Information providing server V Voice Vd Voice data D Voice recognition data C Operation command I Vehicle status information Id Operating status information Ie External information Ii Vehicle side information Il setting Kaijoho N broadcast data Q question message R response message

Claims (8)

A vehicle state information acquisition unit that acquires vehicle state information indicating a state of the vehicle,
An input voice recognition unit that converts input voice into voice recognition data,
On the basis of the voice recognition data, an in-vehicle operation device serving as an operation target device and a voice recognition data analysis unit that determines the operation content for the operation target device,
An operation execution determination unit that determines the validity of executing the operation content on the operation target device based on the vehicle state information,
An operation instruction generation unit that generates an operation instruction for executing the operation content on the operation target device in response to an instruction from the operation execution determination unit;
A device operation unit that executes the operation command;
An output sound generation unit that converts the notification data received from the operation execution determination unit into a sound,
The operation execution determination unit does not execute the operation command when determining that there is no validity, and outputs the notification data for outputting a response message indicating that there is no validity in the output voice. Output to the generator ,
The notification data, together with a response message indicating that there is no validity, includes a question message regarding the execution of the operation content,
When receiving a response to the question message via the input voice recognition unit, the voice recognition data analysis unit executes the instruction to the operation command generation unit according to the content of the response, or A voice operation device for a vehicle , wherein the operation content is not executed on the operation target device. The vehicle has a sensor for detecting at least one of a driving state of the vehicle or peripheral information of the vehicle as vehicle-side information,
The vehicle state information includes the vehicle side information,
The operation execution determination unit, based on the vehicle state information including the vehicle-side information, to determine the validity of executing the operation content to the operation target device,
If the execution of the operation command for the operation target device is determined to be incompatible with the vehicle-side information associated with the operation target device in light of the function of the operation target device, it is determined that there is no validity. The vehicle voice operation device according to claim 1, wherein when it is determined that the vehicle-side information associated with the operation target device is compatible, the validity is determined. The vehicle has a communication device for communicating with the outside of the vehicle,
The vehicle state information includes external information obtained via the communication device,
The operation execution determination unit, based on the vehicle state information including the external information, determines the validity of executing the operation content to the operation target device,
Based on the external information, within a predetermined time after the execution of the operation command for the operation target device, a prediction is made that the operation command for the operation target device is to be executed on the operation target device in a direction opposite to the operation command. 2. The vehicle voice operation device according to claim 1, wherein it is determined that there is no validity in the case, and the validity is determined if the prediction is not established. 3. The vehicle state information includes operating state information of the vehicle-mounted operation device,
The operation execution determination unit, based on the vehicle state information including the operation state information of the operation target device, to determine the validity of executing the operation content to the operation target device,
When it is determined that the operation state information of the operation target device and the operation content match, it is determined that there is no validity, and the operation state information and the operation content of the operation target device are The vehicle voice operating device according to any one of claims 1 to 3, wherein when it is determined that they do not match, the validity is determined.   In the case where the vehicle state information includes, in addition to the operation state information, other information for determining the validity of executing the operation content on the operation target device, the operation target device Determining the validity based on the operating state information, and further performing the validity determination based on the other information only when the validity is determined. The voice operation device for a vehicle according to claim 4. The in-vehicle operating device is configured to be able to set an operation state in stages,
The operating state information includes setting step information indicating a setting step of the operating state,
The operation execution determination unit, when the operation content for the operation target device is determined to be an operation for relatively changing the setting stage of the operation state, the vehicle state information acquisition unit acquires the The voice operation for a vehicle according to claim 4, wherein the validity of executing the operation content on the operation target device is determined based on the vehicle state information including setting stage information. apparatus. A voice input device into which the voice is input,
A notification device for notifying an output from the output sound generation unit,
Connected to a communication network, the vehicle state information acquisition unit, the input voice recognition unit, the voice recognition data analysis unit, the operation execution determination unit, and a voice operation control device provided with the operation command generation unit,
Disposed on the vehicle, the voice input device, the notification device, and a voice operation input / output device provided with the device operation unit,
A communication interface device that is provided in each of the voice operation control device and the voice operation input / output device and connects the voice operation control device and the voice operation input / output device to be communicable via the communication network; The voice operation device for a vehicle according to any one of claims 1 to 6 , further comprising: The voice input device and the notification device are configured by a mobile communication device capable of performing data communication including a mobile phone, a smartphone, a tablet terminal, a notebook computer, an in-vehicle communication device, and a car navigation system. Item 8. The vehicle voice operation device according to Item 7 .

Images