JPS626845A - Diagnostic apparatus for vehicle - Google Patents

Abstract

PURPOSE:To enable effective diagnostic work in short time by sequentially diagnosing the items which can be answered on the basis of using conditions and the items which can be answered only through actual inspection. CONSTITUTION:The diagnostic apparatus for vehicle 23 is provided with means 11 for storing the fault tree information containing first questions to be answered on the basis of the using conditions of vehicle and second questions to be answered through actual inspection. While means 15 for receiving the symptom information from symptom information input means 13 to predict the fault tree formed with fault tree information is provided. On the basis of causes predicted through said means 15, rearward predicting means 17 will search said information tree sequentially from fist to second questions rearwardly while the question means 19 will question about the rearwardly predicted phenomena to output the prediction results from the output means 21.

Description

DETAILED DESCRIPTION OF THE INVENTION [Technical Field of the Invention] The present invention relates to a vehicle fault diagnosis device that applies a so-called expert system.

[Description of Prior Art] As an example of a conventional vehicle failure diagnosis apparatus to which an expert system is applied, those shown in FIGS. 5 to 7 are considered.

FIG. 5 is a schematic diagram of such a vehicle fault diagnosis device,
FIG. 6 is an explanatory diagram of a fault tree applied to this vehicular fault diagnosing device, and FIG. 7 is a processing flowchart of the inference section included in the vehicular fault diagnosing device.

As shown in FIG. 5, the vehicle fault diagnosis device 1 has a knowledge data section 3 and an inference section 5, and the inference section 5 is connected to an operation section 9 via an interface 7. ing. Generally, the knowledge data section 3 is a database, or
The inference section 5 is constructed of a computer and is capable of communicating with an operation section 9 on the user side via an interface 7.

The knowledge data section 3 stores data for forming the fault tree shown in FIG. The fault tree shown in Figure 6 has three causes for the original symptom A1: B+, B2,
B3 is involved, and as shown in the figure, it is shown that the following causes are involved in order in the lower order. In addition, in the failure tree shown in FIG. 6, it is assumed that the symptom (phenomenon>A2) indicated by a broken line is involved in cause B1.

The inference section 5 searches the data stored in the knowledge data section 3, in other words, uses the fault tree shown in FIG. 6 to infer the cause of the symptoms occurring in the vehicle. However, in this inference method, in addition to forward inference that traces the fault tree from the upper level to the lower level, backward inference is applied that traces the fault tree from the lower level to the upper level, and attempts are made to efficiently find the cause of a given symptom. ing. Note that forward inference is inferred in the form Cause (X Y ), that is, the cause of symptom X is Y, and backward inference is in the form Ru1e
(XY), that is, cause Y brings about phenomenon (symptom) X.

Specifically, the inference method of the inference section 3 is performed as shown in FIG. 7, for example.

In FIG. 7, step 701 shows the second step of setting the original symptoms. This is input via the operation unit 9 on the user side, and here it is assumed that symptom A1 has been selected and input with respect to the fault tree shown in FIG.

Step 703 is the cause Yi(B+) of the input symptom A1.
, B2. B3) is shown. This process is performed using forward inference.

Step 705 is the cause of the strike Yi (B+, 82
.

One cause from among B3) Yo (e.g. B+)
This shows the process of extracting .

Step 707 is the extracted cause Yo (B+)
In other words, it shows the process of finding another symptom (phenomenon) x1 for which the cause Yo(B+) is the cause. Of course, this processing is performed by backward inference. In the example in Figure 6, other symptoms A for cause B1
2 exists, so symptom A2 is found as phenomenon X.

Step 709 shows a process of inquiring about the discovered phenomenon X (A2) as to whether the phenomenon X is actually occurring. Here, in response to this question, the user inputs the presence or absence of phenomenon X via the operation unit 9.

Step 711 shows a process of determining the input result input through the operation unit 9. If it is determined that the phenomenon X (A2) exists, the process proceeds to step 715, and if it is determined that it does not exist, the process proceeds to step 713.

Step 713 is a process performed when it is determined in step 711 that the phenomenon X (A2) has not occurred, and the corresponding Yo (
B+) is removed, and then the process returns to step 705, and from step 705 onward, the process proceeds with inference regarding the next cause Yi (82°83). In addition, here, the Yo
(B+) causes the strike Yi(8+, B2, B
3) was excluded because if cause B1 were the true cause, symptom A2 would inevitably occur as a result, but in reality it was determined in step 711. This is because symptom A2 did not occur.

On the other hand, in step 715, the phenomenon X (A
2) shows the process to be performed when it is determined that 2) has occurred. In this case, the fact that the phenomenon X (A2) has occurred does not necessarily mean that the cause Yo(B+) is the phenomenon X (
Since it can be said that there is a high probability of being the true cause of A2), we will now switch the cause Yo (B+) to a new symptom X, and hereafter, we will explain the next cause Y! for this symptom X (B+). (C+.

C2, C3).

Such a vehicle fault diagnosis device uses backward reasoning in addition to forward reasoning to help efficiently infer the true cause.

However, in the above-mentioned vehicle failure diagnosis device, the questions asked to the user after the inference section performs backward inference cannot be answered unless the questions can be answered based on the usage status of the vehicle and the vehicle is actually inspected. Since the configuration was a mix of items, there was a problem that in order to answer these questions, it would be difficult to diagnose the failure unless someone who had knowledge of the usage conditions and someone who could inspect the equipment were available at any given time. .

For example, in the case where a vehicle dealer uses a vehicle failure diagnosis device, the front desk clerk who deals with customers is separated from the maintenance clerk who inspects and repairs the vehicle. Therefore,
Although the front desk staff can answer questions according to usage conditions, they cannot answer questions that require inspection, making it difficult to diagnose failures. In addition, although the inspection staff can answer inquiries regarding inspections, they do not answer questions regarding usage conditions, and as a result, it is difficult to diagnose failures here as well.

[Object of the Invention] This invention improves the above-mentioned problems and provides a fault diagnosis device for a vehicle that can easily perform a fault diagnosis even when a person who has knowledge about the usage situation and a person who can inspect the vehicle are not present at the same time. The purpose is to

[Summary of the Invention] In order to achieve the above-mentioned problem, this invention, as shown in FIG. a fault tree information storage means for storing fault tree information having second inquiry information that can be answered; a symptom information input means for inputting symptom information; and a forward inference means for forwardly inferring a fault tree formed by the fault tree information. and backward inference means for making a backward inference by searching the fault tree in the order of the first inquiry information and then the second inquiry information based on the cause inferred by the forward inference means, and inquiring about the phenomenon for which the backward inference was made. The vehicle failure diagnosis device 23 is configured by comprising an interview means for performing the above, and an inference result output means for outputting the inference result, and the vehicle fault diagnosis device 23 is configured to include questions to be asked to a person who has knowledge of the usage situation and questions to be asked to a person who performs the inspection. It has been made possible to do this separately.

[Explanation of Examples] The present invention will be described in detail below.

FIG. 2 shows a schematic diagram of the vehicle failure diagnosis 111i device.

Although this device 23a does not have the same appearance as the device 1 described in FIG. 5, it is functionally constructed with each of the means shown in FIG. In addition, the first rEI examination information and the second
The interview information may be stored separately in two storage devices, or may be stored in one storage device, divided into areas. Any form is acceptable as long as it is possible.

The main task of the forward inference means 15 is to perform forward inference on the fault tree in the form of Cause (X Y ), that is, the cause of symptom X is Y.

The backward inference means 17 mainly uses Ru1e(XY)
The main task is to make backward inferences in the form of , that is, cause Y brings about phenomenon (symptom) X. Then, the backward inference means 17 performs the inference order first in the first inquiry information and then in the second inquiry information, as will be described in detail in FIG. Note that the backward inference performed within the first interview information is Ru1.
e1. The backward inference performed within the second medical interview information will be referred to as Ru1e2.

The fault tree shown in Figure 3 is [for the engine stall JA+ symptom, ``ignition system j B+, rAAC valve'' B2, ``
It indicates that the cause is "lower fuel pressure" B3, ... "crank angle sensor" B4, and for "ignition system JB+", the phenomenon of "restart NGJ" is caused by rAAC valve J.
For the B2, the phenomenon of ``engine stalling when lightly stepping on the accelerator'' occurs, and for the B3, the phenomenon of ``low fuel pressure'' occurs due to ``insufficient output.''
This shows that a phenomenon called A4 is brought about. or,
Regarding "ignition system" B1, the following cause "spark plug J
C.I.

It shows that the ignition coil JC2 is the cause, and that the cause of the spark plug J'C+ is that it causes the phenomenon of plug abnormality JAS.

Here, the truth or falsehood of the phenomena of ``restart NGJA2, engine stalls when r lightly stepping on the accelerator'' A3, and ``insufficient output JA4'' can be immediately confirmed from the usage status of the vehicle, but [
Regarding plug abnormality JAs, the authenticity cannot be confirmed unless the plug is actually removed and inspected, so Ru1e
(A2, B+) T: The data to be inferred backwards is stored in the failure tree information storage means 1 as the first inquiry information shown in FIG.
It is stored in 1.

Similarly, Ru1e (A3.82), Rtlle
The data retrospectively inferred in (A4, B3) is also stored in the fault tree information storage means 11 as first inquiry information. On the other hand, Ru1e (C
I, A5) [The data regarding the plug abnormality JAS cannot be confirmed to be true or false unless it is actually inspected, so it is stored in the failure tree information storage means 11 as the second inquiry information.
is stored in

The inference processing performed by the inference unit 27 shown in FIG. 2 will be explained using the flowchart in FIG. 4.

Step 401 shows the step of setting the original symptoms.

Now, suppose that ``Entos JAz'' is set as the original symptom.

Step 403 shows the process performed by the forward inference means 15. Here, the symptom [stall JA+ is substituted for X in Cause (X Y ), and the cause Yi of this symptom A+ (B+, B2, 83, . . . , B4).

Step 405 shows a process of selecting a cause Yli to which Ru1e1 can be applied from the cause list Y1. Here, the cause list Yi is the cause Yl to which Ru+ei can be applied.
In addition to i, Y2'% average Hini Ru to which Ru1e2 can be applied
Reason for not being able to apply both 1e 1 and Ru1e 2 Y3
1. Therefore, in step 405, among these causes Yi, causes Y+ i to which Ru181 is applicable are selected.
Select all (B+, B2, B3).

Step 407 shows the application process of Ru1e1.

This process is performed by the backward reasoning means 17.

Here, this application of RLl181 is performed for each cause Y11 selected in step 405, and first, for this one cause Y+o (for example, B1), the phenomenon X+o(
For example, A2 ) is Ru1e (A2 B+ ) and M1
It will be discussed.

Step 409 is inferred by the backward inference means 17 and the phenomenon
This is to ask the diagnostician whether or not the phenomenon of IO is actually occurring. Under the dealer situation shown in the conventional example,
In this case, the person diagnosing the problem is a front-line person. Ru
Since all the phenomena prepared in 1e1 are assumed to be answered without inspecting the vehicle, the diagnostician can immediately answer these questions.

Step 411 shows processing for determining the result of the inquiry in step 409. If the diagnostician answers that the phenomenon x10 exists (YES) in response to the question asked in step 409, the process proceeds to step 415; if the diagnostician answers that it does not exist (NO), the process proceeds to step 413.

Step 413 shows the process of excluding cause Y+o(B+) from the cause candidates since the phenomenon Xl0(A2) does not exist. Step 415 is cause Y because the symptom (A2) is present.
This shows processing for leaving +o (B+) as a cause candidate.

Step 417 indicates a process of determining whether backward inference has been performed for all causes Y+i to which Ru1e1 can be applied. If 82, 83, etc. remain, the process returns to step 405, and the same processes as shown in steps 405 to 413 are repeated until it is determined in step 417 that the process is complete.

Step 419 indicates a transition to the process of searching for the remaining vehicle stalls Y2 i and Y3 i as the cause of the engine stall JA+.

Step 421 shows a process of determining whether or not there is a cause Y21 to which Ru1e2 can be applied in the cause list. If it is determined that there is no cause Y2 i to which Ru1e2 can be applied, the process moves to step 423, where other processing, that is, normal processing that is also performed in the conventional example, is performed.

The normal process is, for example, a process such as displaying a list of cause candidates.

On the other hand, if there is a cause Y2i to which Ru1e 2 can be applied in step 421, for example, if there is a phenomenon called "plug abnormality JAs" in the spark plug JC+ shown in FIG. ．Proceed to the following process.

Step 425 is a cause list Y to which Ru1e2 can be applied.
2 i shows the process of selecting one cause Y20 (for example, C1).

Step 427 concerns the selected cause Y20(CI);
It shows the process of backward inference of the phenomenon X2 (A5) that satisfies Ru1e2.

Step 429 shows a process of asking questions regarding the retrospectively inferred phenomenon x 2 (A5) and prompting inspection processing. The diagnostician, e.g. maintenance, answers this question and proceeds to step 431.
Enter the inspection results using .

Step 433 is to judge whether the symptom X20 is true or false based on the inspection result, and if the inspection result is normal, that is, the phenomenon
If it is determined that (A5) does not appear, it is determined that this cause Y20 should be excluded from the causes, and step 42 is performed.
Return to 1. Also, the inspection result is abnormal, that is, the phenomenon X20
If it is determined that (A5) appears, step 4
Move to 35.

Step 435 is the cause Y20 to which Ru1e2 is applied.
Determine whether there are any further causes below.

Then, if there is another cause, the process moves to step 439,
If not, the process moves to step 437.

Step 437 indicates processing for diagnosing the cause Y20 as the true cause. That is, since cause C1 has no next cause below it and phenomenon A5 appears, it is determined that at least this cause CI is sufficient as one cause of the failure.

Step 439 is a process when it is determined that the following cause exists in the cause Y20, and here, Ca
In order to apply use(XY), the cause Y20 is input to the forward inference means 15 as a new symptom X.

Step 441 shows the processing of the forward reasoning means 15 to search the cause list Yi of the input symptom X.

As described above, this vehicle fault diagnosis device answers questions that can be answered based on the usage status of the vehicle, and then searches for the true cause while actually inspecting the vehicle. This allows the person in charge and the maintenance person to operate separately. Also, the - person's diagnostician first,
It is possible to diagnose the cause of the failure by answering questions that can be answered by TMm based on the usage status of the vehicle, and then performing an inspection. In this case, questions can be asked sequentially, separating items that can be answered based on the usage situation and items that cannot be answered without actual inspection, so the diagnostician can quickly answer questions according to the usage situation, and then... Since these interviews allow for targeted inspections, efficient diagnostic work is possible.

[Effects of the Invention] With the vehicle failure diagnosis device according to the present invention, questions can be asked sequentially, dividing the questions into questions that can be answered based on usage conditions and those that cannot be answered without actual inspection. All you have to do is answer the questions and then perform a targeted inspection, making it possible to carry out efficient diagnostic work in a short period of time.

In addition, in cases where the front desk staff and maintenance staff are working separately, the front desk staff will first respond according to the usage situation.
Next, maintenance staff can perform efficient inspections and use the system in accordance with the actual situation.

[Brief explanation of the drawing]

FIG. 1 is a diagram corresponding to claims of the present invention, FIG. 2 is a schematic diagram of a vehicle failure diagnosis device according to an embodiment of the present invention, and FIG.
Figure 4 is an explanatory diagram of the fault tree used to explain the embodiment, Figure 4 is a flowchart of the process of the inference section shown in the ÷ example of the present invention, and Figure 5 is a schematic diagram of a conventional vehicle fault diagnosis device. , FIG. 6 is an explanatory diagram of a fault tree used to explain the prior art, and FIG. 7 is a processing flowchart of the conventional inference section. 11...
-Fault tree information storage means 13...symptom information input means 15...forward inference means 17...backward inference means 19...interview output means 21...inference result output means

Claims (1)

[Claims] A failure tree information storage means for storing failure tree information having first inquiry information that can be answered according to the usage status of the vehicle and second inquiry information that can be answered by actually performing an inspection; and symptom information for inputting symptom information. input means; forward inference means for forward inferring a fault tree formed by the fault tree information; and based on the cause inferred by the forward inference means, the fault tree is inferred from the first interview information and then the second interview information. A fault diagnosis for a vehicle comprising: backward inference means for searching information in order and making backward inferences; interrogation means for inquiring about the backwardly inferred phenomena; and inference result output means for outputting inference results. Device.