JPS59136641A - Physicochemical apparatus - Google Patents

Abstract

PURPOSE:To improve easiness and adequateness of operation by using a speech recognizer as an input means. CONSTITUTION:When a switch 6 is thrown on, an operatable item and the number thereof are displayed on a display device 5. When an operator inputs a desired number by a speech through a microphone 1 by observing the display, the speech is recognized by a speech recognizer 2 and the recognized number is displayed on a panel by a CPU4. If the display is correct, the respective parts 7, 8 forming a unuclear magnetic resonator, etc. are controlled via the recognizer 2, the CPU4, etc. according to the voice ''yes'' from the microphone 1. The need for operating pushbuttons, switches, etc. is thus eliminated and the operation is made easy. The generation of a malfunction, etc. is prevented and the easiness and adequateness of a physicochemical apparatus are improved.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Application of the Invention] The present invention relates to physical and chemical equipment, and relates to means suitable for use in operating the equipment.

[Prior art]

The physical and chemical equipment is operated by operating switches and knobs. Since these operations are performed manually, 1. There is a possibility of operational errors. 2. Conditions other than optimal values may be set intentionally or inadvertently. For example, in examples such as NM 几 (nuclear magnetic resonance apparatus Goreikoi/L, Atayal setting value, Y, B
It is usually desirable not to change the terms other than the above and the higher-order correction terms. Although it is hidden behind a hidden panel, there is a possibility that it may be changed due to careless operation.5 3.Recently, efforts have been made to automate operations and reduce the number of knobs and switches. . For this purpose, a single switch or knob is used to perform operations for multiple purposes, and certain switches and knobs are overused, resulting in more accidents such as poor contact and breakage than before. .

4. Operators with hand disabilities may not be able to operate the device, or there may be times when the inventor was unable to use his dominant arm, and had difficulty operating the device.

There are drawbacks such as.

[Purpose of the invention]

An object of the present invention is to use a voice recognition device as an input means for inputting operation commands for operating the device, thereby improving ease of operation.
It improves appropriateness.

[Embodiments of the invention]

FIG. 1 shows an example of application to an NMR apparatus. When an operation is started using the switch 6, the CPU 4 displays operable items through the display device 5.

Next, looking at this displayed content, the operator inputs '1' if it is the desired number (for example, 1) from among the numbers added at the beginning of the displayed content. The voice recognition device 2 is
If it is confirmed that the voice is 1,
31, which is expressed as 1 in ASCII code, is transferred to the CPU 4 via the pass line. At this time, 110 equipment 5, ]
From 10 and Y (high), N (no), NG (no)
Light emitting diodes numbered 1 are provided on the panel surface, and the CPU 4 sends a command to light up the diode named 1. Next, the operator confirms that the desired number 1 has been selected using the layer 1. By saying “hai”, CP
U4 executes the command.

If the lighting of the selected diode is incorrect, ′
"Yes," and then specify the correct number again.

If you say “No”, the command will not be carried out, and “
It also has a function that allows you to return to the previous operating procedure specification by commanding ``Mukou''.

FIG. 2 shows one row of speech recognition devices. Audio signal 1,
Data is converted into digital values by an A/D converter and sampled through a microphone, a preamplifier, and a 4KH2 low-pass filter, and is produced in the audio analysis circuit 10. Next, the match (degree of match) with the digital signal read out from the memory 15 in which the feature parameter pattern registered in advance is stored is checked by the match measure calculating circuit 14. The feature parameters are processed as time-series signals, pattern-matched between the time-series feature patterns of the input audio signal and previously registered feature patterns, and the pattern with the highest degree of similarity is selected, and the matching scale calculation circuit 13 selects the pattern with the highest degree of similarity. Output from The time axis matching conversion circuit 14 is a circuit that non-linearly expands and contracts the time lengths of an input audio signal pattern and a registered standard audio signal pattern, thereby temporally matching them on a matching scale. In the determination circuit 17, temporally,
A determination circuit 17 determines which of the combined signals and the standard pattern best corresponds to the voice that is considered to correspond. For example, if it is 1, the ASCII code 31 is output. Speaker voice registration is performed in advance using voice registration mode 16.

According to one embodiment of the present invention, the following effects can be achieved.

1. Misoperation can be prevented compared to push buttons, light bends, and keyboard operations.

2. Compared to push buttons and switches, there are no moving parts at all, improving the reliability of the device.

3. In TEM and SEM, micro-movement of the sample is performed manually.
Instructions are given to change the acceleration voltage and magnification by voice. In NMR, when setting the Golay current among 12 sets, a specific Golay coil is specified by voice, and the current value setting is manually adjusted using one polyurethane coil without a stopper that can be rotated 360 degrees.
You can specify the corresponding Golay coil by voice without taking your hand off the volume control. As in these two examples, operations can be performed simultaneously using both hands or one hand.

4. Remote control becomes possible. When handling clinical test samples,
Alternatively, when working inside the reactor pot area or when handling radioactive isotope samples, if an automatic sample introduction device is used, the operator can operate from a distance from the sample.
After setting the sample on the device, the device can be operated outside the room.

5. Researchers with physical disabilities (handicap);
It can also open the door to operators to operate physical and chemical equipment.

6. There are currently two ways to create standard speech patterns: specific speaker recognition and non-specific speaker recognition. There are usually operating parts that most operators do not need to move. These normally operated parts are generally operated only by the system administrator. In this case, by registering the voice of the person in charge of equipment management as a specific speaker, you can clearly separate the parts that can only be operated by a specific person from the parts that can be operated by unspecified operators. be able to.

For example, in NMR, adjustment of the higher order terms of the Golay coil is preferably operated only by the equipment administrator.

As a modification, as explained during the effect, there is an example in which a voice recognition device is used in combination with a knob, polyme, etc.

In addition, as means for responding from the device, there are light emitting diodes on the panel, displays on the cathode ray tube, etc., but when used in combination with a voice synthesizer, it is possible to interact with the device by voice for both commands and responses.

To output synthesized speech by speech synthesis, it is sufficient to add a circuit that specifies addresses in the symbol memory corresponding to specific speech from the standard speech signal memory and outputs them in a fixed order.

Brief description of the drawing ′
FIG. 1 is a block diagram of an embodiment of the present invention.

Claims (1)

[Claims] l. A physical and chemical device for observation or analysis, characterized in that it is operated via a voice recognition device.