JPS62175060A - Automatic trouble diagnosing system of electronic exchange - Google Patents

Abstract

PURPOSE:To restore trouble only by working as instructed by the system and make better diagnosing by starting trouble diagnosing with a trouble detecting means and by using an inferring means, a test selecting means and a test executing means readily until the trouble is restored. CONSTITUTION:Knowledge of symptoms, kowledge of apparatus constituting an exchange and knowledge of tests are described in a diagnosis knowledge base 4 as classes and a diagnosis working memory 5 expresses them as instances of knowledge. A trouble detecting section 6 supervises messages generated at irregular intervals by an electronic exchange 1, detects trouble and accumulates in the diagnosis working memory 5. An inferring section 7 infers a suspected failed apparatus and a test selecting section 8 selects the most effective method of test for deciding the extent of suspected failed apparatus or for demarcation, and a test executing section 9 issues test command automatically and instructs change of suspected failed apparatus to a person 2 in change of maintenance. The system operates the inferring section 7, test selecting section 8 and text executing section 9 repeatedly until the trouble is restored.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to fault diagnosis of electronic switching equipment, and more particularly to an automatic system for detecting and diagnosing faults in electronic switching equipment and allowing maintenance personnel to repair faulty equipment. .

[Conventional technology]

Conventionally, in fault diagnosis of electronic exchanges, maintenance personnel analyze fault information provided by the electronic switch and subscribers, etc., and follow procedures in the maintenance manual or judgment of maintenance experts until the fault is repaired. Diagnosis was performed by maintenance personnel by repeating numerous tests and replacing suspect devices.

[Problem that the invention seeks to solve]

The conventional method described above has the following problems.

Electronic switching equipment is a very large-scale system, and there are limits to the number of faults that can be diagnosed using the maintenance manual.There are cases in which the fault cannot be repaired even if the procedure described in the maintenance manual is followed. However, it is physically and economically difficult to secure a maintenance expert at all times. Furthermore, although there is always a need to improve maintenance manuals, it is usually not easy to change or add procedures to maintenance manuals.

[Means for solving problems]

The present invention is an automatic system that has the same knowledge as a maintenance specialist and performs fault diagnosis in place of the maintenance specialist, and detects faults by analyzing information provided by electronic exchanges and maintenance personnel. A detection means, a diagnostic knowledge base that describes the maintenance expert's knowledge in fault diagnosis, a diagnostic working memory that expresses the current diagnostic situation, a means for inferring a suspected faulty device, and a test selection means for its verification. and automatic or manual test execution means, the fault detection means starts fault diagnosis, and the inference means, test selection means, and test execution means are repeatedly performed until the fault is repaired.

In the present invention, knowledge of symptoms, knowledge of devices constituting an exchange, and knowledge of tests are described as classes in the diagnostic knowledge base, and the diagnostic working memory is expressed as instances of the above knowledge. Classes and instances are an appropriate expression for organizing similar concepts; classes represent conceptual frameworks, and instances represent entities. For example, if flowers are a class, cherips and tanbobos are instances. In addition, the knowledge of test selection and test result judgment that is not clearly determined is written in a diagnostic knowledge base (described as a rule consisting of multiple condition parts and action parts) so that it can be easily changed and added. This example shows that it is easier to change and add more information than traditional procedural knowledge.

Figure 2 shows an example of procedural knowledge, and Figure 3 shows a corresponding example of rule-based knowledge.
The action parts of the rules corresponding to C, D, E, and F are A' and B.
', C', D', E', and F'. In the example of procedural knowledge, the conditions to which C is applied cannot be determined until all the control flows shown in bold lines are analyzed, whereas in the case of knowledge of rules, the condition part is With a simple control structure that executes the action part of the rule that matches the environment, it can be said that controlling c F is easy because changes to its condition parts cc and c' are localized. . Also, in the case of rule-based knowledge,
Knowledge can also be added locally by creating a condition part that matches the application conditions.

〔Example〕

The present invention will be explained with reference to the drawings.

FIG. 1 shows an embodiment of the present invention, and the system of the present invention includes a diagnostic knowledge base 4, a diagnostic working memory 5, a fault detection section 6, and an electronic exchange 1, a maintenance person 2, and an editor 3.
The control section 10 includes an inference section 7, a test selection section 8, and a test execution section 9.

Diagnostic knowledge base 4 describes the knowledge of maintenance experts for electronic switching equipment, and expresses knowledge of symptoms and knowledge of tests as classes, and knowledge of test selection includes the distribution of suspect parts in the condition part, and the knowledge of tests in the operation part. The knowledge for determining the test result is expressed as a rule that describes the test to be selected in the condition part and the suspect part in the action part. The editor 3 is a maintainer or maintenance expert, and edits the diagnostic knowledge base 4 so that the system can provide better diagnostic results.
can be easily changed or added.

The diagnostic working memory 5 stores fault conditions and test results up to now, and is expressed as an instance of a class in the diagnostic knowledge base.

The fault detection unit 6 monitors the messages that the electronic exchange 1 generates non-regularly, detects faults, stores them in the diagnostic working memory 5, and determines the fault status by checking the degree of urgency and performing statistical processing. , notify maintenance person 2 of the occurrence of a failure,
Start fault diagnosis. It is also possible for the maintenance person 2 to input information.

The inference section 7 infers the suspected faulty device from the fault conditions and test results recorded in the diagnostic working memory 5, and the test selection section 8 selects the most effective method for determining the range of the suspected faulty device or for isolating the suspected faulty device. After selecting a test method, the test execution unit 9 executes the test by automatically issuing a test command to the electronic exchange 1 and instructing the maintenance person 2 to replace the suspected faulty device. is stored in the diagnostic working memory 5. Then, until the fault is repaired, the system operates as follows: inference section 7, test selection section 8, test execution section 9.
Execute repeatedly.

Further, the inference unit 7 selects a rule from among the test result judgment rules, and the test selection unit 8 selects a rule having a condition part that matches the contents of the diagnostic working memory 5, which is the current environment, from among the test selection rules, and determines its operation. Diagnosis can proceed by executing the section.

〔Effect of the invention〕

As explained above, by using the system of the present invention, maintenance personnel can repair failures simply by following the instructions of the system, and further improve the diagnostic knowledge base. This has the effect of making it easier to make better diagnoses.

[Brief explanation of drawings]

FIG. 1 is a diagram showing the overall configuration of the system of the present invention. 1...Electronic switchboard, 2...Maintenance person, 3
...Editor, 4...Diagnostic knowledge pace,
5...Diagnostic working memory, 6...
Failure detection unit, 7...Inference unit, 8...Test selection unit, 9...Test execution unit, 10...
...control unit, Figure 2 is an example of procedural knowledge. A, B, C, D, E, F...Procedure figure 3 is as follows:
This is an example of rule-based knowledge corresponding to FIG. 2.

Claims (1)

[Claims] A fault detection means that detects faults from information provided by electronic exchanges and maintenance personnel, a diagnostic knowledge base that describes the knowledge of maintenance experts in fault diagnosis, a diagnostic working memory that indicates the current diagnostic status, and a diagnostic working memory that describes the current diagnostic status. The coagulation fault device has an inference means, a test selection means for its verification, and an automatic or manual test execution means, and the fault detection means starts fault diagnosis, and the inference means and the fault detection means start the fault diagnosis until the fault is repaired. ,
An automatic fault diagnosis system for electronic switching equipment characterized by repeatedly performing test selection means and test execution means.