JP2002351754A - Input noise discriminating method and its recording medium - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an input noise discriminating method which discriminates noise superposed on an input signal. SOLUTION: By the input noise discriminating method which decides the noise superposed on the input signal from the input signal inputted to a microcomputer, long-range noise which can not be removed by a filter circuit that the microcomputer is equipped with is detected by software and abnormality of the input signal is discriminated as period abnormality or the long-range noise. The long-range noise discriminating means detects whether the input signal is a normal input by regarding an edge of the input signal as an interruption factor and reading in the input signal port level which is a fixed time later by software, thereby detecting the noise.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an input noise determining method, and more particularly, to an input noise determining method for processing an input signal to a microcomputer.

[0002]

2. Description of the Related Art Conventionally, as a method for discriminating noise of an input signal to a microcomputer, a noise filter circuit of a short range in hardware is mounted on the microcomputer itself, and the noise filter circuit detects noise of an input signal to the microcomputer. Had been determined.

Such a prior art is disclosed in, for example,
It is disclosed in JP-A-7-0555437.

In the case of a microcomputer used as an in-vehicle ECU, a method of monitoring a watch dog pulse cycle by a plurality of microcomputers and an integrated circuit is well known as a method of detecting runaway of the microcomputer (FIG. 6). reference).

[0005] When a short-time noise pulse is input due to an abnormal period of the watch dog pulse period or the mixing of external or internal noise of the microcomputer, it is also an important task to separate these noise pulses.

When the input signal has periodicity as noise (input signal abnormality) caused by software accompanying the software, the cycle is monitored by the edge detection function of the microcomputer. Recognizes that the input signal is abnormal.

As a specific example, as shown in FIG. 5, a normal or abnormal input signal is determined after a lapse of time from a previous edge input by using a software interrupt by edge detection of a microcomputer (step S1). S51).

Then, when the operation is normal, the fail counter is cleared (step S53). In addition, the failure counter is counted up at the time of abnormality (step S52).

[0009]

However, the conventional input noise determination method has the following problems.

That is, in the above-mentioned prior art, the first problem is that the microcomputer does not detect abnormality in the noise filter function mounted on the microcomputer. This is due to the configuration of the noise filter circuit in which input signals shorter than the set time are simply discarded.

As a second problem, this function cannot be set at an arbitrary time, and the versatility of the microcomputer causes a problem.
It is generally applied only to noise in an extremely short range such as electrostatic noise, and there is a change in set values during system development, so that it may not be practical to incorporate it as a dedicated circuit.

A third problem is that the input signal abnormality determination method using software, which is a conventional technique, is limited to input signals. That is, in this method, the input signal must have periodicity. In the case of an input signal having no periodicity, a normal range cannot be defined and normal diagnosis cannot be performed.

A fourth problem is that even when the above-mentioned conventional method is used with a periodic input signal, an interrupt is generated even if a noise of several μs level is input. Even if it is the same as the period abnormality, it is detected as a failure.

An object of the present invention is to provide an input noise discriminating method for discriminating such short-range noise as noise by software and discriminating it as an abnormality other than a period abnormality.

[0015]

An input noise determination method according to the present invention is an input noise determination method for determining, from an input signal input to a microcomputer, noise superimposed on the input signal. A long-range noise discriminating method is provided in which a long-range noise that cannot be completely removed by a filter circuit provided is detected by software, and an abnormality of the input signal is discriminated into a period abnormality and a long-range noise.

Further, the long-range noise discrimination method of the input noise discrimination method of the present invention has a configuration in which an edge of the input signal is used as an interrupt factor and noise is discriminated by an interrupt process of the software.

In the long-range noise determining method of the present invention, the edge of the input signal is used as an interrupt factor, and the input signal port level after a predetermined time is read by the software, whereby the normal operation is performed. The long-range noise determination method determines whether or not the input is an input and detects the noise by reading the input signal port level after a predetermined time by software using an edge of the input signal as an interrupt factor. In the case of noise, when the noise is detected based on the input signal port level, the number of times of the noise is counted to determine the system abnormality.

[0018]

Embodiments of the present invention will be described below in detail with reference to the drawings.

FIG. 1 is a flowchart of an input noise determination method according to the first embodiment of the present invention.

Referring to FIG. 1, the input noise discriminating method according to the first embodiment of the present invention comprises only a programmable microcomputer (which can detect an input pulse edge and can use a software interrupt).

In the interrupt processing, the level of the input port used for the interrupt is read, and it is determined whether or not the level is the active level (step S11).

At this time, the determination position is adjusted based on the minimum input pulse length (pulse input time) of the input pulse.
This can be realized by adding a weight or the like.

Here, if a normal level (low level (hereinafter referred to as Lo level) becomes active in the case of an interrupt due to a falling edge), a normal input signal is determined if it is a Lo level at the timing of this determination. judge.

Here, when the signal is not at the active level but at a high level (hereinafter, referred to as Hi level), it is an input signal which should not exist originally, and can be regarded as noise.

Next, the fail counter 2 is counted up (step S12). Since it is noise, the cycle determination is not performed.

On the other hand, if the input level is the active level (Lo level), the number of normal inputs is counted, and it is determined whether or not normal input (non-noise input) is detected continuously (step S13).

If it is continuous, the fail counter 2 is cleared (step S14). This corresponds to a delay at the time of normal recognition with respect to the input frequency of noise, and the noise detection rate (toughness) for noise can be changed by increasing or decreasing the set value of the number of continuous detections.

Then, since it is not noise, the same cycle determination as that of the conventional input noise determination method is performed (step S15).

The above is the details of the processing contents in the present invention. Thereafter, the failure counter counted as described above is determined at an appropriate place such as a background, and an abnormality detection process is performed.

FIG. 3 shows the operation of these fail counters in the form of a waveform. Input signal is active at Lo level, interrupt factor is falling edge, fail counter 2 clear condition (continuous normal detection count setting value)
Shall be 5 times.

At the noise input timing, the fail counter 2 counts up (10), and the cycle determination is not performed, so that the fail counter 1 does not change.

Then, in a normal cycle abnormality (no noise input), only the fail counter 1 is counted up (11). The input at this time is illegal as a cycle, but is not a noise input, so the number of continuous normal detections is counted.

Finally, the fail counter 1 is cleared by the normal cycle input and the fail counter 2 is cleared because the non-noise fall has continued five times (12).

A series of software flows is a repetition of this operation. By monitoring the number of times of these counters, a system abnormality can be detected.

Next, the conditions that can be detected when the present invention is applied will be described.

FIG. 4 shows the timing of the input signal Lo active, the cycle determination when the interrupt factor is the falling edge, and the noise detection timing.

The points indicated by points A and B in FIG. 4 are noise. The point A can be detected because the rising occurs immediately after the interruption occurs, but the interruption does not occur at the point B because the rising does not occur.

That is, since the noise is determined by monitoring the port level immediately after the occurrence of the interrupt,
When the interrupt factor is one edge, a direction that cannot be detected due to noise occurs.

The same applies to the case where this is set to the rising edge. In this case, B can be detected as noise and A cannot be detected.

These problems can be avoided by setting the interrupt factor to both edges. By setting both edges, the falling of both points A and B,
Since an interrupt occurs at each rising edge, it is possible to detect each of them.

Therefore, it is necessary to use different designs depending on the use of the input signal according to each system.

As described above, the detectable conditions when the present invention is applied are summarized as follows: (1) The output time of the input signal must be always longer than the time required to execute the port level determination. That is, the output time of the input signal is always longer than the interrupt generation time and the processing time in the interrupt. (2) The direction of detectable noise is determined by software settings, and is effective for noise immediately after the occurrence of an edge detection interrupt.

It should be noted that the present invention is not limited to the above embodiments, and it is clear that the embodiments can be appropriately modified within the scope of the technical idea of the present invention.

Further, the above-described processing of the input noise determination method can be recorded on a recording medium as a program for causing a computer to execute the processing.

[0045]

As described above, according to the present invention, as a first effect, when a long-range noise that cannot be removed by a noise filter of a microcomputer is applied to an input signal, a failure caused by a conventional input cycle is caused. , Can be detected by software as abnormalities due to noise.

The second effect is that the present invention is applicable to signals other than periodic input signals. For example, when software reset processing is performed by a reset input signal, when noise of several μs
The reset processing was executed by software and the input signal was erroneously detected.However, when the present invention is applied, it is possible to determine the noise as noise. It can be prevented before it happens.

[Brief description of the drawings]

FIG. 1 is a flowchart of an input noise determination method according to a first embodiment of the present invention.

FIG. 2 is a diagram showing the timing of software interrupt occurrence and determination for an input signal.

FIG. 3 is a time chart showing the operation of the processing shown in FIG. 2 with respect to the input signal waveform.

FIG. 4 is a time chart showing a timing at which a cycle determination and a noise detection are performed in the processing shown in FIG. 2 with respect to an input signal waveform.

FIG. 5 is a flowchart of a conventional general diagnosis process.

FIG. 6 is a block diagram of an example of a microcomputer applied to an in-vehicle ECU.

[Explanation of symbols]

 3 Interrupt generation 4 Judgment 10 Count of fail counter 2 11 Count of fail counter 1 12 Number of times of normal detection t1, t3, t7, t11 Period judgment t5 Noise detection t8 Noise not detected S11 to S17 Step

Claims (8)

[Claims] An input noise determining method for determining, from an input signal input to a microcomputer, noise superimposed on the input signal, a long-range noise that cannot be completely removed by a filter circuit provided in the microcomputer. And a long-range noise discriminating method for detecting an abnormality of the input signal into a period abnormality and a long-range noise by detecting software. 2. The input noise discrimination method according to claim 1, wherein the long-range noise discrimination method uses an edge of the input signal as an interrupt factor and performs noise discrimination by an interrupt process of the software. 3. The method of claim 3, wherein the edge of the input signal is used as an interrupt factor, and the input signal port level after a predetermined time is read by the software to determine whether the input is normal. 3. The input noise determination method according to claim 2, wherein noise is detected. 4. The long-range noise discriminating method uses the edge of the input signal as an interrupt factor, reads the input signal port level after a predetermined time by software, discriminates noise. 4. The input noise determination method according to claim 3, wherein when noise is detected based on the input signal port level, the number of times of the noise is counted to determine a system abnormality. 5. The long-range noise discrimination method, wherein the edge of the input signal is used as an interrupt factor, and the input signal port level after a predetermined time is read by software to discriminate the noise. 5. The input noise discrimination method according to claim 4, wherein the normality of the cycle is detected based on the input signal port level, and the number of normal noises in the cycle is counted to determine a system abnormality. 6. The input noise determination method according to claim 3, wherein said long range noise determination method sets a noise detection threshold. 7. The input noise determination method according to claim 3, wherein said long range noise determination method sets a normal recognition delay. 8. A recording medium on which a program for causing a computer to execute the processing of the input noise determination method according to claim 1 is recorded.