JPS60198595A - Voice message output method - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

DETAILED DESCRIPTION OF THE INVENTION Technical Field The present invention relates to a method for outputting a voice message to a man-machine interface of a data processing device.

Various messages are transmitted from the data processing device to the operator through the man-machine interface.

One means of message transmission is audio transmission, which has a different usefulness from visual transmission means such as displays.

When the data processing device performs I-process control, the messages sent to the overerator include /process alarms, sequence messages, operation confirmation messages, and alarm summaries.These messages indicate when a corresponding event has occurred. will be communicated to the operator each time.

Events that require a message occur randomly, with multiple events occurring one after the other within a short period of time, or occurring almost simultaneously.In contrast, with voice messages, by their nature, only one message is spoken. Since it takes a certain amount of time to output one message, a request to output another message may occur while one message is being spoken. At this time, a request that occurs later may have a higher priority, so in order to appropriately deal with such a case, it is necessary to output a message according to the priority.

Purpose> An object of the present invention is to provide a voice message output method that can process output requests according to priority.

<Main points> The present invention specifies the output method for each voice message, either by interrupting the preceding message and outputting it, or by outputting it after the preceding message has finished outputting. In addition, a unique number is determined according to the output priority order, and when a voice message output request occurs, the requested message is output in such a way that the output of the preceding message ends. If the output is rejected and the output method is to interrupt the preceding message and output it, the priority is determined based on the unique number of the requested message and the unique number of the preceding message, and whether output is possible or not. The above object has been achieved by a voice message output method that determines the voice message output method.

<Example> Hereinafter, the present invention will be explained in detail with reference to Examples.

FIG. 1 is a conceptual block diagram of an example of a data processing device to which the present invention is applied, FIG. 2 is a conceptual diagram of a voice message output device in the device of FIG. 1, and FIGS. 3 to 10 teeth,
FIG. 1 is a flow diagram of a program according to an embodiment of the present invention.

<Configuration> In FIG. 1, 11 to 1o are process control units, and 2 is an operator operation unit. A plurality of process control units 11~
1n is connected to the operator operation unit 2 via a communication line or the like, and constitutes one data processing device for process control as a whole. The process control units 11 to 1n all include the C1 computer and the process i! It consists of input/output devices for IIJ tel, and uses computer functions to perform process control within its own range and data processing related thereto.

The operator operation section 2 also consists of an input/output device for input and human-machine communication, and uses the function of the y1 computer to process data for monitoring/operating the man-machine communication and process control sections 11 to 1n of the A operator. Do the following. Input/output devices for man-machine communication in the operator control unit include a voice message output device in addition to a display and a keyboard.

A conceptual diagram of the voice message output device is shown in FIG. Second
In the figure, 21 is a voice generator, 22 is an 81 that controls it.
A perfect program, 23.24.25 is program 22
This is a group of data used by. Data group 23.2
4.25 exists in the computer's main memory.

Of these, 23 is a data group related to a message that needs to be output newly, including data specifying how to output the message, a unique number indicating the type of message, data specifying a task to be activated after the message is uttered, and syllable information of the message. The way to output a message is to interrupt the message currently being uttered and output a new message TN'? There are two predetermined methods: one in which the message is uttered, and one in which the message is uttered after the message currently being uttered is finished, and one of these is specified by data. The output method also includes the number of times the message is uttered. The number of utterances may be infinite, three times, one time, etc., and the number of times to be uttered is determined according to the degree of importance.

The unique scent not only indicates the type of message, but also indicates the output priority. Here, the lower the number, the higher the priority.

24 is a data group regarding messages for which the output request has been accepted ulJ6, and has the same structure as the data group 23.

Reference numeral 25 indicates flags, which include an output service flag, an accepted interrupt wait flag, and a voiced interrupt wait flag.

The output samurai flag indicates the presence or absence of a message registered in the output samurai state, and the accepted interrupt flag indicates the state in which the syllable information from the voice generator 21 has been accepted. The voiced interrupt flag indicates a state in which there is a voiced interrupt from the voice generator 21.

The detailed structure of the program 22 is shown in FIGS. 3 to 10. Here, FIG. 3 shows a first routine for outputting a message, and FIG. 4 shows a second routine for outputting a message. The first routine (message output) is activated when outputting a single message, and is activated when outputting a process alarm, sequence message, operation confirmation message, etc.・Ru.

On the other hand, the second message output routine is activated when multiple messages are output in succession, and multiple messages corresponding to multiple alarms are output in order, such as in an alarm summary. It is activated when

According to the first routine for message output shown in FIG. 3, when activated due to occurrence of a process alarm, etc., in step 1, the output method and unique number are determined based on the contents of the alarm. Next, in step 2, a voice message output check subroutine (details will be described later), it is checked whether a message can be output. If the output is rejected, the process branches to step 6, where the invoking task is activated by a timer and the process ends; however, if the output is possible, the process branches to step 1.
In step 1i3, syllable information for the message is prepared, in step 1114, the output method and fixed i number are determined as in step 1, and in step 5
Executes the voice message output subroutine (details described later). As a result, if the output of the message is accepted, the process ends, but if it is rejected (step 6 is processed and the process ends).

According to the second message output routine in FIG. 4, when activated by an alarm summary output request from the operator, the same processing as the routine in FIG. 3 is performed from step 1 to end. At step 61, alarm summaries and the like are aborted, and between steps 4 and 5, settings are made to start the own task after the occurrence of step 41. This is a measure for outputting the second message uR of alarm 9. As a result, when the first message of the alarm summary has been output, this routine is started again, and in step 7 the output method and unique number are set, and in step 8 the routine is checked to see if the voice message can be output. Execute. As a result of executing this subroutine, 4
A split in the street occurs. That is, possible (match), possible (mismatch)
, J [no noodles (match), and rejection (no match). Tsukasa (
If it is a match), the process branches to step 3, and if it is acceptable (a mismatch) or rejected (a mismatch), the process branches to step 61, which is rejected (a match).
If so, the process branches to stage 111i9. Step 9 is self-task activation processing using a timer. Note that the term "match" or "mismatch" here refers to the match or mismatch of the unique numbers.

Note that this voice output check subroutine is performed in the third
Although this is the same subroutine as Step 2 in FIG. 4 and FIG. 4, Step 2 in both figures is expressed in two branches: accept and reject, regardless of whether the unique numbers match or do not match.

Both Figure 5 and Figure ff16 are message mute routines, and of these! The routine shown in Figure T5 is activated when the message output event is canceled.
The routine shown in FIG. 6 is activated in response to the operator's key operation for muting the sound.

According to the routine shown in FIG. 5, a voice message output permission check subroutine is executed in step 1, and in this case, two branches are performed: match and non-match, regardless of whether the message is accepted or rejected. If there is no match, the process ends, but if there is a match, data for message muting is set in step 2, and in the voice message output subroutine of step 3,
Silence processing is performed. If the mute process is accepted, the process ends, but if it is rejected, the process branches to step 4, where the own task activation process is performed using a timer, and the process ends.

According to the routine shown in FIG. 6, data for muting is set in step 1, and a voice message output subroutine is executed in step 2. As a result of the execution of this subroutine, even if the muting process is accepted and C is also rejected, the process ends.

FIG. 7 shows an interrupt processing routine, which is activated by an accepted interrupt and a voiced interrupt given from the voice generator 21. When the voice generator 21 receives the syllable information of the message to be uttered, the voice generator 21 receives]! The message is returned as 1 [1 gram], and when the message has been uttered a predetermined number of times according to the received voice information, the uttered interrupt is returned.

According to the routine of FIG. 7, when there is an accepted interrupt, the accepted interrupt holding flag is reset in stage 11i3, the output subroutine in stage 4 is executed, and the process ends. When there is an interrupt that has been voiced, the task is activated after the voice has been voiced in step 1, the voiced interrupt flag is reset in step 2, the accepted interrupt flag is reset in step 3, and the output is output in step 4. Execute the subroutine and exit.

8 to 10 are subroutines used in each of the above routines, and FIG. 8 is an output subroutine;
FIG. 9 is a voice message output subroutine, and FIG. 10 is a voice message output check subroutine.

According to the output leave routine shown in FIG. 8, in step 1 it is determined whether there is an output, and if there is no output, the routine returns, but if there is an output, the routine branches to step 2. At step 2, it is determined whether or not there is an accepted interrupt. If there is an accepted interrupt, the process returns; otherwise, the process branches to step 3. Step 3: It is determined whether or not there is a uttered interrupt. If there is, the process branches to step 4, and if there is no interrupt, the process branches to step 5. In step 4, the method of output is determined, and if it is determined that the message should be uttered after the previous message has been uttered, the process returns; otherwise, the process branches to step 5. In step 5, the sound output device (voice generator) is activated. At this time, syllable information corresponding to the content of the message is given to the voice generator. Next, stage N6
The output flag is reset, the accepted interrupt wait flag is set, and the voiced interrupt flag is set, and the process returns.

According to the voice message output subroutine shown in FIG. 9, step 1r: It is determined whether or not there is an output samurai, and if there is one, the process branches to stage N2, and if there is no one, the process branches to stage 4. In step 2, the method of output is determined, and if it is determined that the message should be uttered after the previous message has been uttered, the process branches to a rejection return, otherwise the process branches to step 3. Step rP33TELL Compare the unique number of the new IJ2's message with the fixed number of the message registered before it, and if the number is not small, branch to a rejection return, but if it is small, step Branches into 4. In step 4, the new message is registered as a message to be output, and the output flag is hit. Then, at stage N5, the output subroutine is executed and the process returns to reception.

According to the voice message output check subroutine shown in FIG. 10, it is determined in step 11J1 whether there is an output or not, and if there is, the process branches to step 2, and if there is no output, the process branches to step 4. At step M2, the method of output is determined, and if it is determined that the message should be output after the preceding message has been generated, the process branches to step 5; otherwise, the process branches to step 3. In stage N3, the unique number of the new message is compared with the unique number of the previously registered message, and if the unique number of the new message is not smaller than the registered unique number, the process branches to step 5; branches to stage 4. In step 4, the unique number of the new message and the registered unique number are compared, and if they match, the process branches to an acceptable return (unique number matches), and if they do not match, it branches to an acceptable return (unique number 1 does not match). do. Note that these returns correspond to OK (match) and OK (mismatch), respectively, of the four branches at step 8 in FIG. At stage 111i5, the unique number of the new message and the registered unique number are compared, and if they match, the process branches to a rejection return (unique number match), and if they do not match, the process branches to a rejection return (unique number does not match).

These correspond to rejection (match) and rejection (mismatch) of the four branches described above, respectively.

<Operation> The voice message output method configured like a child allows
The operation when outputting a voice message is as follows.

<Single message output operation> First, the case of outputting a single message will be explained. Most messages are single messages,
Process alarms, sequence messages, operation confirmation messages, etc. are all output as a single message.

These messages are prioritized for output in this order. Note that although the message is a single message, it may be uttered an infinite number of times or a predetermined number of times depending on its importance.

First, a message output operation based on a single request will be explained as the simplest example of message output. In addition,
It is assumed that the preceding message output operation has already been completed.

When a message output request occurs, the routine shown in FIG. 3 is executed. First of all, Dan lol! In 11, how to output and the fixed f
j I @ is set, and in step 2, a voice message output checkable check is executed.

The details of the voice message output check subroutine are shown in FIG. 10. If there is no message with output, the process branches from step 1 to step 4, where the unique number of the message and the preceding message that has already been output are displayed. Compare the unique numbers of the messages.

The comparison results in a match or mismatch, but either of these results in an acceptable return.

Therefore, audio information is prepared in stage N3 of FIG.
The output method and unique number are set again in step N5, and the voice message output subroutine is executed in step N5.

The details of the voice message output subroutine are shown in FIG. 9, where it branches from step #li1 to step 4, registers the message as a message with output, and sets the output flag. Then, at stage N5, an output subroutine is executed.

The details of the output subroutine are shown in FIG. 8, and the process branches from step 1 to step 2 to determine whether there is an accepted interrupt. Since the preceding message has already been output, there is no hold,
The process branches to step 13 to determine whether there is a voiced interrupt.

Since the preceding message has already been output, there is nothing to do, and the process branches to step 5, where the voice output device R (voice generator) is activated.

Then, in step 6, the output flag is reset, the accepted interrupt flag is set, and the voiced interrupt flag is set, and the process returns.

When the output cup routine returns, the voice message output subroutine becomes an acceptance return, and as a result,
The routine of FIG. 3 ends.

Through the above-described operations, message voice information and the like are given to the voice generator, and the message voice output operation is activated.

When the voice generator receives the given voice information, etc., it returns an accepted interrupt signal to the program and starts uttering the voice message.

When the accepted interrupt signal is returned, the program resets the accepted interrupt flag in step 3 of the interrupt handling routine shown in FIG. 7, and executes the output subroutine in step 4. Then, the output subroutine shown in FIG. 8 returns from stage rli1, and the interrupt processing routine ends. That is, in response to the accepted interrupt, the accepted interrupt flag is reset. Thereafter, when the voice generator finishes uttering the message audio output, a uttered interrupt is returned, and the uttered interrupt holding flag is reset by the corresponding interrupt processing routine. This completes the message output operation based on a single request.

Next, as a somewhat complicated message output operation, a new message output request occurs while the output of the preceding message is activated, and then a new message with a higher priority is output. Let us explain what happens when a request occurs.

Suppose now that audio information of a certain message is given to a voice generator to activate audio output, but the voice generator has not yet returned an accepted interrupt. In this state, the output flag is reset, and the accepted interrupt flag and voiced interrupt flag are set.

In this state, for example, if a sequence message output request occurs, in step 1 of the routine in Figure 3, the output method is set to interrupt the previous message and output the unique number. For example, it is set to 3. Then, in step 2, a check is performed to determine whether or not a voice message can be output.

In the voice message output capability check subroutine, step 1 branches to step 4, from which a match or non-match acceptability return is made.

Then, in step 3 of FIG. 3, voice information etc. for the sequence message are prepared, in step 4 the output method and unique number are set again, and in step 5 the voice message output command is executed.

In the voice message output subroutine, step 1 branches to step 4, where the sequence message is registered as a message to be output and an output flag is set. Then run the C1 output routine.

In the output subroutine, it is determined in step 1 whether the output is present, and the process branches to step 2. Then, the process returns because the C1 acceptance interrupt flag is set. This return becomes an acceptance return in the audio message output υ Brugen, and the routine shown in FIG. 3 ends.

Through the above operations, the output flag remains set, and the sequence message is registered in the output status.

If a process alarm occurs in this state, the routine shown in FIG. 3 is executed again, and in step 1, the method of outputting a process alarm message and the setting of a unique number are performed. The method of output is to interrupt the preceding message and output, and the unique number is set to 1, for example. Next, a subroutine for checking whether a voice message can be output is executed.

In the voice message output possibility check subroutine, the process branches from step 1 to step 2, and in step 2 the method of output is determined, and if the method is to interrupt the preceding message and output, the process branches to step 3. In step 3, the unique number 1 of the process alarm is compared with the unique number 3 of the sequence message registered in the state with output, and as a result, the process branches to step 4. Then, from step 4, the reception of unique number mismatch becomes 1 ■ function return.

Next, in stage rPi3 of FIG. 3, the voice information for the wrestling alarm etc. is set as #(#L), the output method and unique number are set in stage 4, and the voice message output rib routine is executed in stage 5.

The voice message output subroutine passes through stages 1 and 2 and reaches stage 3. In stage 11J3, the unique number 1 of the process alarm is compared with the unique number 3 of the sequence message registered before it, and as a result, stage 1@i
The process branches to step 4, and in step 4, registers a process alarm and sets an output wait flag. As a result, a process alarm is registered instead of a sequence message. Then, in step 5, the output subroutine is executed.

In the output subroutine, step 1 branches to step 2,
It is determined whether there is an accepted interrupt.

Assuming that the voice generator has not yet returned an accepted interrupt for a message whose output was activated prior to the gistance message, the program returns here and the routine of FIG. 3 ends.

Thereafter, when the voice generator returns an accepted interrupt, the interrupt handling routine is executed, the accepted interrupt flag is reset, and the output interrupt is executed. In the output subroutine, the process goes through step 1.2 and reaches step 3, where it is determined whether or not there is a voiced interrupt. Until the audio output of the preceding message is finished, the process branches to step 4 because it is in a holding state, and the method of outputting the process alarm registered in the output waiting state is determined, and it interrupts the preceding message. Since it is the method of outputting the audio, the process branches to step 5 and the audio output device is activated. As a result, audio information for process alarms and the like is given to the voice generator. That is, the output of the process alarm with the highest priority interrupts the message being outputted before it, and is activated in priority to the sequence message that has been output. Then, at stage N6, reset the output flag,
Sets the accepted interrupt flag and generated interrupt flag and returns. Returning at this time marks the end of the interrupt processing routine.

After that, in response to the accepted interrupt from the voice generator, receive 1! I!
The expressed interrupt flag is reset, and the generated interrupt flag is reset in response to the voiced interrupt.

Note that if the process alarm at this time is a heavy alarm and is to be repeatedly uttered an infinite number of times, the utterance will continue indefinitely and no uttered interrupt will occur.

To stop the generation of such messages, a mute operation by the operator is required.

<Output operation of multiple messages> Next, the output operation of multiple messages will be explained. An example of outputting multiple messages is outputting an alarm summary or the like.

Assume that the preceding message has now been uttered, and that an alarm summary output request is generated in this state.

The routine shown in FIG. 4 is executed in accordance with the alarm summary output request.

In step 1 of FIG. 4, the method of outputting the alarm summary and the setting of a unique number are performed. The output 4 Hijikata is to be output after the preceding message is uttered, and the unique number is set to 5, for example. Then, in step 2, a subroutine for checking whether voice messages and messages can be output is executed.

In the voice message output possibility check subroutine, from stage 1 to stage 11i! Branch to ii4, and there 1. Return acceptable for X, 0, and 0 number matches or mismatches.

Next, the syllable information of the first message of the alarm summary c1 is prepared in step 3 in FIG. is set, and the voice message output subroutine is executed in step 5.

In the voice message output subroutine, the program branches from step 1 to step 4 to register the alarm summary and set the output flag, and in step 5, executes the output rib routine.

In the output subroutine, step 1.2 leads to step 3, from which it branches to step 5 to activate the audio output device. As a result, the voice generator is given audio information such as the first message of the alarm summary. Then, in step 6, the output holding flag is set, the accepted interrupt holding flag and the voiced interrupt holding flag are set, and the process returns. This return is the acceptance 4 return of the voice message output rib routine, and the routine of FIG. 4 is ended.

Eventually, the voice generator returns an accepted interrupt, and the accepted interrupt flag is reset by the interrupt processing routine.

After that, when the voiced interrupt returns, the interrupt handling routine is executed again, starting the task after voiced in step 1, and then passing through step 2.3.4 to set the voiced interrupt flag and accept. The flag for holding interrupts is reset and the process ends.

Then, step 1 in Figure 4 first! l! Since i41 is set to launch its own task after utterance, 4 meetings where utterance has already been made
A Mu + k 4-t Hibiki-"□1□-1+ + -
+ Inya J (ffi ho 1self i +J-f ÷
1 move.

By the way, if a message output request with the highest priority occurs while the voice generator is activated for the alarm summary and is waiting for an accepted interrupt or has a voiced interrupt, that request will be given priority. , since the alarm summary output operation is interrupted, the accepted interrupt or voiced interrupt that has just been returned is not necessarily the alarm summary. However, in any case, the routine of FIG. 4 is activated upon the first returned voiced interrupt.

When the routine shown in FIG. 4 is started, in step 7, the method of outputting the second message of the alarm summary and the setting of the unique number are performed. The output method and unique number of the second message are the same as the first message. Then step 8C
Execute the checker flow to check if voice message output is possible.

Step 1 of voice message output check subroutine
Then, the process branches to stage 2 or stage 4 depending on whether there is an output or not.

For the m group shunted to stage R144, it is determined whether it matches or does not match with the fixed number 1 recorded in a record, and a possible match or mismatch is returned depending on each case.

In the case of a possible match return, since the preceding message that has been uttered is the first message of the alarm summary, the process branches to stage 1II3 in FIG. Performs message output processing. For subsequent messages, output processing is performed in the same way as long as possible matches continue.In the case of possible returns that do not match, the process branches to step 61 in Figure 4 and the alarm summary is aborted. Let's do it. This means that another message with a higher priority was output between the first message and the second message of the alarm summary, so in order to avoid confusing the message listeners, This will discontinue the output of subsequent alarm summaries.

If the voice message output possibility check subroutine branches to step 2, it branches to step 5 to determine whether the unique @ sign matches/mismatches, and returns a rejection of match or mismatch depending on each case. do.

In the case of a match rejection return, the process branches to step 9 in FIG. 4, where the invoking task is activated by the timer; in the case of a non-match return, the process branches to step 61 in FIG. 4, where the alarm summary is also triggered. Make the cut.

In the manner described above, a plurality of messages such as alarm summaries are output as audio.

<Message Mute Operation> Mute the message being uttered is performed by the mute routine shown in FIGS. 5 and 6. These routines are used to mute the sound in accordance with the disappearance of the reason for message output or the operator's key operation when the number of message output repetitions is infinite.

For example, when the cause of the process alarm disappears, the mute routine shown in FIG. 5 is activated, and in step 1 a voice message output check subroutine is executed. The method of muting is determined by the method of interrupting and outputting the preceding message, and the unique number is made equal to the unique number of the alarm message that is being uttered.

In the voice message output possibility check subroutine, it is determined in stage ii whether or not there is an output, and the process branches to either stage 2 or 4.

When branching to stage 2, it passes through stage 3 and reaches stage 5,
It determines whether the unique number is accepted or not, and returns a rejection if it matches or does not match.

When the process branches to stage N4 and is IC, it is determined whether or not the unique numbers match, and a possible match or mismatch return is executed.

If the unique numbers do not match, the process in FIG. 5 ends regardless of whether acceptance is possible or rejection is possible. If the unique f1 number does not match, it means that the registered message and the cause of muting do not correspond.
No sound is muted.

If the unique numbers match, regardless of whether acceptance is possible or refusal, in Figure 5, the first child L to be split into stage 2
At T mbeme 9, Rikuta's Dardamasuya is performed, and at step 3, the voice message output subroutine is executed.

In the voice message output subroutine, it is determined in step 1 whether there is an output or not, and the process branches to step 2 or 4.

The command that branches to stage 2 makes a rejection return through stage l@3. When a rejected return is made, the self-task is activated by a timer at step M4 in FIG. 5, and the process ends.

When branching to step 4, data for muting is registered, an output flag is set, and an output subroutine is executed in step 5.

In the output subroutine, step 5 is reached through steps 1.2.3.4, and data for silencing is given to the audio output #ik position to activate it. Then, in step 6, the output wait flag is reset, the accepted interrupt flag and the generated interrupt flag are set, and the process returns. As a result, the voice message output subroutine accepts and returns, and the mute routine ends.

When the audio/J device accepts the data for muting and returns a τ mating completed interrupt, the interrupt processing routine resets the accepted interrupt flag, and when the muting is completed, the interrupt processing routine returns the voiced interrupt. The voiced interrupt flag is reset by

Muffling by key operation is performed by setting mute data in step 1 and executing a voice message output subroutine in step N2, as shown in FIG. The routine shown in FIG. 6 is simplified from the routine shown in FIG. 5 by omitting the process of checking whether a voice message can be output or not and the process of activating the own task using a timer.

Effects> As described above, according to the present invention, it is possible to realize a voice message output method that can process output requests according to priority order.

[Brief explanation of the drawing]

FIG. 1 is a conceptual diagram of an example of a computer processing device to which the present invention is applied; FIG. 2 is a conceptual diagram of a part of the audio output device; FIGS. 3 to 10 are examples of the present invention. 11-1n...Process control section 2...Operator operation section 21...Generator 22...Programs 23-25...Data group FIG. Figure 8 Figure 9 Figure 1θ

Claims (1)

[Claims] For each voice message, one of two ways to output it is determined: either the preceding message is interrupted and outputted, or the preceding message is outputted after the output ends, and the method is determined according to the output priority. Article IIIB has been established, and when a request to output a voice message occurs, if the requested message is to be output while waiting for the output of the preceding message to finish, the output will be refused, and the output will be rejected. If the method is to interrupt the preceding message and output it, the voice that determines the priority and determines whether to output it based on the unique number 1) of the requested message and the unique number of the preceding message. Message output method.