JP3082847B2 - Test method and system for voice response device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for testing a voice response apparatus that edits and outputs a combination of various fixed voices registered in a memory.

[0002]

2. Description of the Related Art There is a voice response device which edits a date, time, quantity, numeral, and other various fixed voices in combination in accordance with an instruction from a host device and outputs the synthesized voice. Conventionally, in this type of voice response device, a hardware test such as a memory test or a highway test is performed by transmitting and receiving test data when the service is stopped.
This is done by the maintenance person confirming the sound output from the speaker.

For example, Japanese Patent Application Laid-Open No. 57-142700 discloses a voice response device provided with a test data generation circuit and a test data detection circuit, and inputs low-frequency test data outside a voice band into a voice output unvoiced section. A voice response device monitoring method is described in which the test data is detected by a test data detection circuit to monitor the voice response device for abnormality or the like.

In Japanese Patent Application Laid-Open No. 58-116597, a voice processing unit is provided with a voice response section test circuit and an output voice designation switch. The test circuit is activated, and the output voice is controlled by the switch. A description is given of means for making a designation, outputting a synthesized sound from a speaker, and checking whether sound is normally output.

[0005]

SUMMARY OF THE INVENTION The above-mentioned Japanese Patent Application Laid-Open No. 57-14 / 1982
In the voice response device monitoring method described in Japanese Patent Publication No. 2700, edited voice is not used as test data.
There is a problem that not only the hardware of the voice response device but also a function test including voice editing cannot be tested.

On the other hand, in the audio processing apparatus test system described in Japanese Patent Application Laid-Open No. 58-116597, a test of the audio editing function can be performed. However, when the test is performed, the service must be stopped. In addition, it is necessary to confirm the edited voice with a speaker, and it is difficult to automate the test.

[0007] It is an object of the present invention to provide a large-scale voice response system in which a plurality of voice response devices are lined up, the voice memory being normal, the time-division multiplexing circuit being normal, and the voice editing processing being normal. An object of the present invention is to provide a voice response apparatus test method and system capable of confirming a transmission function.

[0008]

SUMMARY OF THE INVENTION A method and system for testing a large-scale voice response apparatus according to the present invention include a plurality of voice response apparatuses each of which is time-division multiplexed and output to a specified test time slot. The audio digital signal is dropped and stored in the memory of the test apparatus, and the test apparatus determines the beginning of the audio by detecting a silent section of the audio digital signal read from the memory, and compares the data dropped from the time slot. The normality of the audio data, the normality of the time division multiplexing circuit, and the normality of the audio editing process are determined.

More specifically, in a large-scale voice response system in which a plurality of voice response devices are connected to a device having a switch function, for example, an exchange, an exchange control unit controls the plurality of voice response devices with respect to the plurality of voice response devices. A voice type is specified, an editing method, and a vacant time slot are designated, and a voice transmission instruction is issued, and an exchange is instructed to multiplex the time slot on a highway connected to the test apparatus. The test apparatus drops data of a plurality of time slots to be input and stores the data in an audio memory provided in the test apparatus. Then, the head of the voice is detected by detecting a silent section of the stored voice data, and the normality of the operation of each voice response device is confirmed by comparing a plurality of dropped data.

[0010]

FIG. 1 is a block diagram showing a first embodiment of the present invention, and shows a large-scale voice response system in which a plurality of voice response devices are connected to an exchange. In FIG. 1, reference numerals 101-1 to 101-4 denote voice response devices,
102 is a voice memory, 103 is a voice control unit, 104 is a voice response unit, 105 is an exchange, 106 is an exchange control unit, 10
Reference numeral 7 denotes a test apparatus, 108 denotes an audio memory, 109 denotes a test control unit, and 110 denotes a separation circuit.

Voice data is registered in advance in each voice response device 101, and voice is managed by the number of a voice piece. The voice specified by the voice output instruction from the exchange control unit 106 is output to the specified time slot.

The voice response devices 101-1 to 101-4 have a voice memory 102 in which various voice data are registered.
And a voice control unit 103 that performs audio editing processing for an arbitrary time slot in accordance with an instruction from the exchange control unit 106.
And an audio response unit 104 that accumulates edited audio, multiplexes the audio on the audio highway, and outputs the multiplexed audio.

With such a configuration, the voice response device 10
1-1 to 101-4 time-division multiplex an arbitrary sound into an arbitrary time slot and output it. This voice response device 10
A large-scale voice response system is realized by connecting 1-1 to 101-4 with the exchange 105.

The exchange control unit 106 designates information necessary for arbitrary voice editing and a time slot to a plurality of voice response devices 101-1 to 101-4, and issues a voice transmission instruction. Is instructed to multiplex the time slot on the highway connected to the test apparatus 107.

In the test apparatus 107, data of a plurality of input time slots is dropped by the separation circuit 110 and stored in the audio memory 108. The test control unit 109 detects the head of the voice by detecting a silent section of the stored voice data, and compares the plurality of dropped data to determine the normality of the operation of the voice response devices 101-1 to 101-4. Can be confirmed.

Next, the operation of the present invention will be described with reference to FIG. Usually, voice response devices 101-1 to 101
In the case of -4, when there is no voice output instruction from the exchange control unit 106, silence data is output to all time slots.

When an event for requesting voice response to the voice response devices 101-1 to 101-4 occurs, the exchange control unit 106
Designates information necessary for voice editing and a time slot to the voice response devices 101-1 to 101-4 to instruct voice output. Each voice response device 101-1 to 101-4
Realizes a multiplex audio output function by outputting edited audio to a designated time slot.

Next, in the present system, for the operation when testing the voice response apparatus, for example, a test of the transmission operation of voice n as information necessary for voice editing is performed, and the voice response apparatuses 101-1 to 101-4 are tested. The following describes an example in which the above-described operations are performed on the four devices.

The exchange control unit 106 includes the voice response device 101
Specify an empty time slot a for -1 and
The voice output of the voice n is instructed, and the vacant time slot b is designated to the voice response device 101-2, and the voice n
, And designates a vacant time slot c to the voice response device 101-3, instructs a voice output of the voice n, and specifies a vacant time slot to the voice response device 101-4. By designating d, the audio output of the audio n is instructed.

Each of the voice response devices 101-1 to 10-10
In 1-4, the voice control unit 103 stores the voice memory 1
The corresponding voice data is read out from 02, the voice n is edited, multiplexed on the voice highway via the voice response unit 104, and output.

The exchange control unit 106 sets the time slot a of the voice highway of the voice response apparatus 101-1 to the time slot 0 of the highway connected to the test apparatus 107 and the time slot of the voice highway of the voice response apparatus 101-2. b is the time slot 1 of the highway connected to the test apparatus 107, the voice highway time slot c of the voice response apparatus 101-3 is the time slot 2 of the highway connected to the test apparatus 107, and the voice response apparatus 101.
-4 audio highway time slot d
The exchange 105 is instructed to exchange data in the time slot 3 of the highway connected to the network 07, and the test apparatus 107 is instructed to perform data comparison in the time slots 0 to 3.

In the test apparatus 107, each data is stored in the audio memory 108 while the time slots 0 to 3 are dropped by the separation circuit 110, and the head of the audio data is detected while the test control unit 109 searches for silence. I do. After detecting the head of the voice data in each time slot, the data from the voice response devices 101-1 to 101-4 is compared.

The test device 107 detects abnormalities in all functions of the voice response devices 101-1 to 101-4, such as the voice memory, the voice editing process, and the time division multiplexing circuit, based on the data comparison results.

FIG. 2 is a block diagram showing a second embodiment of the present invention. In FIG. 1, each voice response device 1
01-1 to 101-4 and the test device 107 are provided independently, but in the present embodiment, each voice response device has a test function independently.

In FIG. 2, each voice response device 201-1
Each of the voice response devices 201-1 to 201-4 includes a voice memory 202, a voice control / test control unit 203, and a voice response / separation unit 204. Two functions are realized.

When an event requesting voice response to the voice response devices 201-1 to 201-4 occurs, the exchange control unit 106
Specifies the information and time slot necessary for voice editing and instructs the voice response device in which the event requiring voice transmission has occurred to output voice. The designated voice response device outputs the edited voice to the specified time slot. Next, when testing the voice response devices, the exchange control unit 106 sets each of the voice response devices 201-1 to 201-4.
, A vacant time slot is designated, and an audio output of the audio n is instructed. In each of the voice response devices 201-1 to 201-4, the voice control / test control unit 203 reads out the relevant voice data from the voice memory 202, edits the voice n, and edits the voice n.
Output to 4.

Each voice response / separation unit 204 separates the synthesized voice n of the designated time slot and stores it in each voice memory 202. Then, each voice control / test control section reads out the stored synthesized voice n, compares the data, and determines whether or not the voice synthesis function is operating normally.

[0028]

The present invention employs a method of detecting an abnormality by comparing output data multiplexed on the highway after voice editing, so that not only hardware of the voice response device but also voice editing is performed. Testing of the included functions can be performed.

Also, since the system is designed so that a plurality of devices can be tested at the same time, it is not necessary to check the sound with a speaker or the like for each device, it is possible to automatically check the sound, and the test efficiency can be increased.

Further, according to the present invention, since a test is performed using an available time slot online, there is no need to stop a service for the test, and the efficiency of system operation can be improved.

[0031]

[Brief description of the drawings]

FIG. 1 is a block diagram showing a first embodiment of the present invention.

FIG. 2 is a block diagram showing a second embodiment of the present invention.

[Explanation of symbols]

 101-1 to 101-4 Voice Response Devices 102 and 108 Voice Memory 103 Voice Control Unit 104 Voice Response Unit 105 Switch 106 Exchange Control Unit 107 Test Device 109 Test Control Unit 110 Separation Circuits 201-1 to 201-4 Voice Response Device 202 Voice memory 203 Voice control / test control unit 204 Voice response / separation unit

Claims (6)

(57) [Claims] 1. A method for testing a voice response device in which a plurality of voice response devices and one test device are connected to an exchange having an exchange control unit, wherein the exchange control unit is provided in the plurality of voice response devices. The type of audio, the editing method, while specifying an empty time slot and instructing audio transmission, and instructing the exchange to multiplex the time slot on a highway connected to the test apparatus, The test apparatus separates the input data of the plurality of time slots, accumulates the data in an audio memory provided in the test apparatus, and detects a head of audio by detecting a silent section of the accumulated audio data. A method of testing a voice response device, comprising determining a normal operation of the voice response device by comparing the plurality of separated data. 2. A method for testing a voice response apparatus in which a plurality of voice response apparatuses are connected to an exchange having an exchange control unit, wherein each of the voice response apparatuses is assigned to an empty time slot designated by the exchange control unit. A voice response device for transmitting voice data edited by a specified voice type and an editing method, and determining the normality of the voice response device by monitoring the voice data in the empty time slot. Test method. 3. An apparatus comprising: a plurality of voice response devices; a test device; an exchange to which the plurality of voice response devices and the test device are connected; and an exchange control unit; For the answering device,
Specifying different test time slots, instructing to edit and output test audio,
Means for instructing the exchange to multiplex the time slots on a highway connected to the test apparatus, wherein the test apparatus separates the input test audio data of the plurality of time slots. And a means for storing the stored voice data in a voice memory provided in the test apparatus and monitoring the stored voice data. 4. The voice response device by detecting a head of voice by detecting a silent section of voice data for test stored in the voice memory, and comparing the plurality of separated data. 4. The test system for a voice response device according to claim 3, wherein the normality of the operation is determined. 5. An apparatus comprising: a plurality of voice response devices; an exchange to which the plurality of voice response devices are connected; and an exchange control unit, wherein the exchange control unit is provided for the plurality of voice response devices.
A means for designating a test time slot, instructing to edit and output a test voice, and wherein each of the voice response devices according to an instruction of the exchange control unit has a designated empty time slot. And a means for transmitting voice data edited by a specified voice type and editing method, and extracting and monitoring the voice data of the transmitted empty time slot. system. 6. Each of the voice response devices registers various voice data and stores a voice memory for storing the edited test voice data, and edits a voice for an arbitrary time slot according to an instruction from the exchange control unit. While performing processing, a voice control / test control unit having a test function of the test voice data stored in the voice memory, and voice data edited by the voice control / test control unit are multiplexed on a voice highway. 6. A voice response / separation unit for outputting and outputting the test voice data to the voice memory for storage.
A test system for a voice response device according to claim 1.