JP3104320B2 - Vibration waveform discriminating inference 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 vibration waveform discriminating inference apparatus which is attached to, for example, a household gas appliance and is used to automatically stop the gas appliance when an earthquake occurs.

[0002]

2. Description of the Related Art When an earthquake occurs, it is necessary to stop gas appliances in order to prevent a fire from occurring, etc. When there is vibration in the appliances, it is determined whether or not an earthquake has occurred and the gas appliances are automatically stopped. Equipment is attached.

[0003] Earthquake discrimination in such a safety device,
Recently, for example, an oscillating wave is set as an ON / OFF signal by comparing with a set value level, and characteristic quantities such as the number of signals, the maximum value of the ON time, the maximum value of the OFF time, the ON time ratio, and the OFF time ratio are obtained. This is performed by discriminating and inferring the vibration waveform from the feature amount.

[0004]

However, performing the discriminant inference of the vibration waveform from a large number of feature quantities as described above requires a long time to calculate the feature quantities and requires a large amount of program memory for discriminative inference. There was a problem of becoming.

SUMMARY OF THE INVENTION The present invention has been made in view of the above problems, and has as its object to provide a vibration waveform discriminating inference apparatus capable of quickly performing vibration waveform discriminating inference with a small number of discriminating inference program memories.

[0006]

According to the present invention, there is provided an ON / OFF signal output means for outputting an ON / OFF signal corresponding to a vibration waveform.
A measuring means for measuring each of the ON time and the OFF time in the OFF signal; and a counting means for classifying each of the same time zones obtained by the measurement into each time zone and counting each time zone. Means, and the count value of each time zone obtained by the counting means within the set time and divided into each time zone
The relationship between the degree of conformity of the vibration waveform and the desired vibration waveform is determined.
Inference means for inferring vibration waveform discrimination based on the indicated membership function .

[0007]

According to the present invention, ON / OFF corresponding to the vibration waveform
The ON time and the OFF time of the OFF signal in the same time zone are counted, and the respective count values within the set time and the vibration waveforms divided into the respective time zones are shown.
Member indicating the relationship between the desired vibration waveform and the fitness
Vibration waveform discrimination inference is performed based on the ship function .

[0008]

DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described below with reference to the drawings.

FIG. 1 is a block diagram showing an entire configuration of an embodiment of a vibration waveform discriminating and inference apparatus to which the present invention is applied.

[0010] The vibration waveform discriminating and inferring device 1 generates O
A seismic device 2 as an ON / OFF signal output means for outputting an N / OFF signal, and a microcomputer 3;
The microcomputer 3 detects O in the ON / OFF signal.
Measuring means 4 for measuring each of N time and OFF time, counting means 5 for counting the same time period by measurement, time counting means 6 for set time, and counting obtained by counting means 5 within the set time The operation as the inference means 7 for performing the inference of the vibration waveform based on the numerical value is performed.

The above-mentioned seismic device 2 is provided with a movable piece which swings in accordance with the vibration of the device, and comes into contact with a fixed contact when the vibration becomes larger than a certain level. ON signal when contacting fixed contact, O when non-contact
FF signals are respectively output.

Next, the processing operation of the microcomputer will be described with reference to the flowchart of FIG.

First, the seismic sensor 2 is turned on as shown in FIG.
When the / OFF signal is given, the ON time t1, t3, t5, t7, t9... And the OFF time t2, t4, t6, t8... In the ON / OFF signal are measured (step 1). The above t1 ... is 0.1 sec.
To 1.0 sec. The ON / OFF signal is based on a vibration waveform, as shown in FIG. 3, and is turned ON when the vibration level becomes higher than a set value and turned OFF when the vibration level falls below the set value.

The above ON times t1, t3, t
5, t7, t9 ... OFF time t2, t4, t6, t8 ...
Are counted in the same time zone within the set time T (step 2), fuzzy inference is performed based on the counted result, and the inference result is output (step 3,
4). The count value at the time t1...
0.2 sec) is [4] of t4, t5, t7, t8, and the middle time zone (0.3 to 0.6 sec) is t1, t2, t3.
At [4] at t9, a long time zone (from 0.7 sec) becomes [1] at t6.

Next, an example of fuzzy inference based on measured values will be described.

FIG. 4 shows the membership functions (A) and (B) of the antecedent part used for fuzzy inference. The membership function (A) is a membership function for a long time zone. The axis is a goodness of fit, and has three membership functions, large, medium, and small. The membership function (B) is a membership function for a short time zone, and has two membership functions, large and small, with the horizontal axis representing the count value and the vertical axis representing the fitness.

The fuzzy inference rule is composed of the following four rule groups, each of which includes an antecedent part concerning a time zone and a consequent part of an earthquake or impact as an inference result.

(1) If long time zone is large then
Earthquake (2) If long time zone is small then shock (3) If long time zone is medium and short time zone is small
then earthquake (4) If Long time zone is medium and short time zone is large
Then impact The above rules were created in view of the characteristics of the frequency characteristics of the seismic wave and the shock wave.

When the count value is given, the MIN calculation is performed in each antecedent part to calculate the degree of conformity, and the degree of conformity is defined as the degree of conformity of the consequent part to each earthquake or impact. For example, if the conformity of the rule (1) is set to 0.
5, when the fitness of the rule (2) is 0.2, the fitness of the rule (3) is 0.3, and the fitness of the rule (4) is 0.4, the fitness is highest by selecting the maximum value. In this case, 0.5 of rule (1) is selected. In this case, the fuzzy inference result is an earthquake. When the equipment to which the vibration waveform discriminating and inferring device 1 is attached is, for example, a gas appliance, the gas supply valve is closed by an earthquake signal from the device 1.

Then, after the fuzzy inference output is performed,
The count value for the previous time is cleared and the count for the next fuzzy inference is performed (steps 5 and 2). In this way, the inference output is sequentially performed with the passage of time.

[0021]

According to the present invention, the ON / OFF signal corresponding to the vibration waveform is counted in the same time zone of the ON time and the OFF time, and the ON / OFF signal corresponding to the vibration waveform is counted. Each count value and
The vibration waveform divided into each time zone is the desired vibration waveform
Vibration waveform discriminating inference is performed based on the membership function indicating the relationship with the degree of conformity, and discriminating inference is performed based only on the count value of the ON / OFF time in the same time zone. Vibration waveform discriminating inference can be performed quickly by discriminating inference program memory.

[Brief description of the drawings]

FIG. 1 is a configuration diagram of a vibration waveform discrimination inference device according to the present invention.

FIG. 2 is a flowchart for explaining the operation of the present invention.

FIG. 3 is an explanatory diagram of an ON / OFF signal and a vibration waveform.

FIG. 4 is an explanatory diagram of a membership function used for fuzzy inference.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Vibration waveform discrimination inference apparatus 2 Seismic sensor (ON / OFF signal detection means) 4 Measuring means 5 Counting means 7 Inference means

Claims (1)

(57) [Claims] An ON / OFF signal output means for outputting an ON / OFF signal corresponding to a vibration waveform; and an ON time and an OF of the ON / OFF signal.
A measuring means for measuring the respective F Time (4), those of the same time period by the measuring divides each time slot
A counting means (5) for counting each time zone; and a count value of each time zone obtained by the counting means (5) within a set time and a vibration waveform corresponding to each time zone are desired.
Membership that shows the relationship between
An inference means (7) for performing a vibration waveform discriminating inference based on a loop function .