JPS63204118A - Testing apparatus - Google Patents

Abstract

PURPOSE:To automate the diagnosis of test data and to increase a working speed, by providing a data diagnostic apparatus in which a trouble mode effect analytical sheet forming means, a diagnostic route forming means and a diagnostic means are incorporated. CONSTITUTION:The test data from an artificial satellite 1 is subjected to data processing/editing by the software test data processing means 11 of a data processing apparatus 3 through a sub-system testing apparatus 2 to be inputted to a data diagnostic means 4 and planning data is put to practical use by a trouble mode effect analytical sheet forming means 12 incorporated as software to perform talk processing and a diagnostic route is automatically formed by a diagnostic route forming means 13. Next, diagnostic processing is performed on the basis of the judge reference rule of a knowledge data base part 15 and the algorithm of an inference function part 16 and it is judged whether the test data is normal to automatically display an abnormal data measure.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a test device for testing and confirming the function and performance of an artificial satellite as a device under test on the ground, for example.

[Conventional technology]

Conventional satellite testing equipment can be roughly divided into subsystem testing equipment that tests each subsystem of the satellite, and data processing equipment that processes data from this subsystem testing equipment, as shown in Figure 8. It consisted of. In the figure, (1) shows the satellite as the device under test that undergoes function and performance tests on the ground, and (2) shows the satellite's subsystems, such as the power supply system that supplies power to each part of the satellite, and its attitude.・Subsystem testing equipment that tests each subsystem such as the attitude and orbit control system that performs orbit control, (3)
is the above data processing device.

The subsystem test equipment (2) tests the functions of the satellite's subsystems, namely the power supply subsystem, telemetry/command subsystem, attitude/orbit control subsystem, propulsion subsystem, and thermal control subsystem. This is to test on the ground whether the performance is as designed.

For example, in a functional test of the attitude/orbit control subsystem,
In order to perform posture control as designed, it is tested whether the posture detection sensor, posture control electric circuit, posture control drive mechanism, etc. operate normally. .

Furthermore, the performance test will test whether the attitude control accuracy and control speed are within the design value range.

All of the test output data for each subsystem described above is sent to the data processing device (3). Data processing device (3
) has built-in test data processing means as software, and processes/edits test data. For example, the output data of the attitude/orbit control system test includes voltage values that indicate the detection sensitivity of the attitude detection sensor,
The above includes analog data such as current values and voltage values for various parts of the electric circuit, as well as binary gaintal data indicating the ON/OFF status of each switch, binary digital data such as rotation speed indicating control speed, etc. The data is input to the data processing device (3), and data processing/editing is performed by the test data processing means. Generally, a general-purpose large-sized computer or a mini-computer is used as the data processing device (3), and all output data of the above test is processed/edited as binary digital data.

The content of data processing/editing in the data processing device (3) is as follows:
Mainly, engineering value conversion processing that converts 2M numerical data into decimal numbers and unit conversion into engineering value units, classifying various data collected at each time and putting it into a table, and graphing how numerical data changes. My main focus is on editing. The data after the data processing/editing is output to a line printer or cathode ray tube.

Satellite experts diagnose the test data while looking at these output results. Diagnosis of test data means determining whether the functions and performance of each subsystem of the satellite are performing as designed, and the design values are used as the criteria for this determination.

Furthermore, from the test data that does not satisfy the design values, the location of the failure, such as where in the subsystem is malfunctioning, is estimated based on past experience and knowledge.

Conventional satellite test equipment is configured as described above, and the diagnosis of the test data of each subsystem of the satellite is done manually.Satellite experts judge whether the test data is good or bad and estimate the location of the failure. It was done. In addition, conventional satellite test equipment includes various design data at the time of design,
For example, what is commonly called ``failure mode and effect analysis'' conducted during design
(Abbreviation r FME J: Failure
Mode and Effect Analysis
There was no built-in means (software) for effectively utilizing results such as is) at the time of testing, and it was done manually, so design data was not efficiently and effectively utilized.

[Problem to be Solved by the Invention 3] As explained above, conventional satellite testing equipment relies on manual diagnosis of a huge amount of test data, requiring a large number of satellite experts, and For data diagnosis, there was a problem in that it took an enormous amount of time for experts to compare test data with design drawings and design data to determine whether the test data was acceptable or not, and to estimate the location of defects.

This invention is intended to solve the above-mentioned problems by connecting a data diagnosis device to a conventional data processing device, effectively utilizing design data at the time of design during testing, and automating the diagnosis of test data. The purpose is to save labor and shorten test time.

Means for Solving Problem C] The test device according to the present invention connects a conventional data processing device to a data diagnosis device, and the data diagnosis device effectively utilizes design data at the time of design for diagnosis at the time of testing. Two types of means devised to make this possible are incorporated as software: failure mode and effect analysis route creation means and diagnostic rule creation means.

Furthermore, in the data diagnosis device described above, an "expert system," which is commonly known in the fields of knowledge engineering and artificial intelligence, is built as software (means) for diagnosing test data. This "expert system" is referred to herein as a diagnostic means and consists of a knowledge database section and an inference function section. The knowledge database section contains judgment criteria for test data diagnosis in the form of rules as a database, and these judgment criteria use design values at the time of designing the artificial satellite as reference values. The inference function part is a software I/ware that incorporates an algorithm for H1-arguing the result using reasoning such as a so-called syllogism, and is a collection of rules (if ... then ... Used when inferring diagnostic results from

The above-mentioned data diagnostic equipment is capable of processing software (programs) written in Lisp language and Prolog language, commonly known as "artificial intelligence languages", at high speed.
Uses what is commonly known as an "inference machine." As an example of an inference machine, a sequential inference machine has been commercialized, sold, and is actually in operation in Japan.

[For production]

In the invention of Party B, the failure mode and effect analysis sheet creation means built into the data diagnostic device is used to predict and analyze the failure mode of the satellite in advance at the time of design, analyze its effects, and develop countermeasures. In addition to being able to
Based on these analysis results, the diagnostic rule creation means can create diagnostic rules (standardized judgment criteria for test data diagnosis) to be used for test data diagnosis during testing.
Analysis results during design can be efficiently used for diagnosis during testing.

Furthermore, the diagnosis of test data is automated by a diagnostic means consisting of a knowledge database section with test data diagnosis rules and an inference function section, and the speed of test data diagnosis processing is increased by connecting a dedicated data diagnosis device. Test time can be shortened.

〔Example〕

FIG. 1 shows the overall configuration of an embodiment of the present invention, focusing on the hardware, and (1) to (3) are the same as the conventional device described above. (4) is a data diagnostic device, which is connected to the conventional data processing device (3). Software is incorporated into the data processing device (3) and data diagnosis device (4) as means, and each process is executed according to an algorithm. The above data processing device (3
) and the software configuration incorporated in the data diagnosis device (4) are shown in FIG.

In this figure, the test data processing means (11) is incorporated into the data processing device (3) and is the same as conventional means.

Next, the failure mode effect analysis sheet creation means (12), the diagnosis rule creation means (13), and the diagnosis means (14) are software built into the data diagnosis device (4), and the above-mentioned diagnosis means (14) Knowledge database department (1
5) and an inference function section (16).

In the satellite test equipment configured as above,
Test data from the satellite is sent to the subsystem test equipment (2
) is input to the data processing device (3), and data processing/editing is performed by the test data processing means (11). The processing up to this point is the same as that of the conventional device described above. The test data processed/edited by the test data processing means (11) is input to the data diagnosis device (4),
The test data is diagnosed by the diagnostic means (14).

By the way, the present invention has been devised to effectively utilize the design data at the time of designing when performing the diagnosis of the test data. In other words, in the present invention, before diagnosing the test data, that is, at the time of design, the failure mode and effect analysis sheet creation means (12) and the diagnosis rule creation means (13) are operated in sequence (8). A means for making effective use of the above test data in diagnosis is incorporated as a software means.

These software means will be explained in detail below. Generally, when designing a satellite, the design! As a result, the various functions of a system or subsystem are usually grouped into functional blocks, and a functional block diagram is created that shows the input/output relationships between each block.

The failure mode and effect analysis sheet creation means (12) performs failure mode and effect analysis (FMEA) based on the functional block diagram created at the time of design, and uses the analysis results as FM.
Software means to compile it as an EA sheet (
program).

As one of the systematic analysis methods, the above FME person
It is an analysis method used during design that is widely used throughout the world.
In this invention, the processing flow of the failure mode and effect analysis sheet creation means (12) is configured based on this analysis method (FMEA method), and the data diagnosis apparatus (12) in which this means (12) is incorporated 4) and the satellite designer (expert) can create the FME person C1- while performing dialogue processing. The dialog processing here refers to, for example, responding to a question displayed on a CRT screen such as a graphic display of the data diagnosis device.
This refers to a method in which an expert enters the answer using a keyboard, etc.

Failure mode and effect analysis sheet) and creation means (12) are FME
It consists of an A format 1 creation program and a question generation processing program, and these formats will be explained in detail below.

(a) FMEA format ll'F configuration program This program is based on the above-mentioned FME eight techniques (IzME
This is a program that draws the format (framework) of a person sheet, and has the same function as the processing flow of a program that draws the framework of a general table. Figure 3 shows the format of the FME person sheet drawn by this program. An example was given. In this figure, each column (21) to (30) is given a name according to the content of the division described by the double dialogue process, but these are simply used in the FMEA method described above. According to the classification adopted.

(b) Question generation processing humulogram This program generates question sentences to fill in the blanks in each column of the FME human track created by the above-mentioned FMEA format creation program, and displays them on the CRT screen. indicate. This is a program that takes in the answers from the experts to the questions and embeds them in each column of the FMEA sheet, and these processing operations have the same functions as a general dialogue processing program. In order to explain the processing operation of this program in more detail, a specific example of question sentences sequentially generated by this program according to the processing is shown in the form of flowchart 1 in FIG. Each question part (41) to (50
), the expert enters the answers from the keyboard of the data diagnosis device (4), and this program embeds the answers entered in each column of the above FME person C1-, and It serves to complete the sheet.

By using the failure mode and effect analysis sheet creation means (12) consisting of the two types of programs (a) and (b) explained above, the creation of an FMEA sheet based on the above FMEA method can be performed with the data diagnosis device (4). This can be done through dialogue processing with experts.

Next, the diagnostic rule creation means (13) uses the FME created by the failure mode effect analysis sheet creation means (12).
This is software means for creating test data diagnosis criteria rules (diagnosis rules), which will be described later, from the contents of the person sheet. The above diagnosis rules are rules contained in the knowledge database section (15) of the diagnosis means (14) described later.
It is a collection of

In order to explain the processing operation of this diagnostic rule creation means (13), as a specific example, from the contents of the FME person sheet of the gas jet device (subsystem) used to control the orbit and attitude of the artificial satellite, The process of creating a diagnostic rule used for diagnosing the test data of this gas jet device using the diagnostic rule creation means (13) will be described below.

FIG. 5 is an example of the FME seat of the gas jet device. In this figure, the content described in the failure mode (24) corresponds to the assumption part in the diagnosis rule described later, and the cause of failure column (25) and the effect column of the failure mode (26) , and the contents described in the countermeasure recommendation column (30) correspond to the conclusion part of the diagnosis rule, which will be described later. For example, in Figure 5, the radial thruster (3
In the case of "external leakage of liquid or gas" (32) among the descriptions in the failure mode column (24) in 1), the diagnosis rule is as shown below.

(Hypothesis part): If external leakage of liquid or gas has occurred from the radial thruster, (Conclusion part): A defective seal or debris clogging has occurred, resulting in loss of manpower and attitude/orbit control.

With double seal structure, ring Leak test before and after the environmental test It is necessary to carry out tests as a countermeasure. be.

As is clear from the above specific diagnosis rule side, the assumption part and conclusion part of the diagnosis rule correspond to the description contents in each column of the above FME person sheet, and based on this correspondence, the assumption part and conclusion part This means that the contents of the diagnosis rules can be automatically created.

FIG. 6 is a flowchart 1 representing the flow of processing for explaining the operation of the diagnostic rule creation means (13), and in this figure, processing (61), (62), (63) ,
(65) is automatically processed by the diagnostic rule creation means (13), but in the process of (64), the diagnostic rule automatically created in the process of steps (61) to (63) is It includes a process of checking the contents and correcting them if necessary by experts, and is finally checked and
The corrected diagnostic rules are registered in the knowledge database section and are effectively used for diagnosing test data.

As is clear from the above explanation, the failure mode effect analysis sheet 1-creation means (12) and the diagnostic rule creation means (13)
), the design data at the time of design can be used to diagnose the test data.

Next, a diagnosis means (14) for diagnosing test data during a test using software similarly incorporated in the data diagnosis device (4) will be described below.

Diagnostic means (14) is commonly known in the fields of knowledge engineering and artificial intelligence as “
It is a rough 1-ware (program, means) called "expert system" and has a knowledge database section (15).
and an inference function section (16), which operates within the data diagnosis device (4) described above and performs diagnostic processing on test data.

In the knowledge database unit (15), criteria for diagnosing test data are converted into rules, and a database is configured as a collection of rules. An example of the test data diagnosis rule is shown in FIG. 7 and will be described below.

In Figure 7, five diagnostic rules are shown from (71) to (
75). Each rule consists of a hypothesis part (76) and a conclusion part (77), and the hypothesis part (76) is "If...
The conclusion part (77) corresponds to "then...". For example, in this figure, the meaning of rule (75) is [If posture detection sensor No. 61 is normal,
Gyro N001 is normal and wheel No.

1 is abnormal, there is a possibility that there is a problem with the wheels of the attitude/orbit subsystem 1. As a countermeasure, check the wheel drive circuit and input signal level.''

As explained above, the collection of these rules is the knowledge database section (15).

Next, the inference function unit (16) is composed of a general algorithm commonly called an "inference engine" in the fields of knowledge engineering and artificial intelligence, and uses reasoning such as so-called syllogism to infer the result. An algorithm has been created in software and is used when inferring a diagnostic result from a collection of the above-mentioned "if...then..." rules.

Excited knowledge database section (15) and inference function section (1
6) is used to diagnose the test data, determine whether it is normal or abnormal based on the test data, and automatically determine and display the location of the problem based on the abnormal data and the countermeasure at that time.

By the way, in the above explanation, the case where this invention is applied to a testing device for an artificial satellite has been described, but it is applicable not only to an artificial satellite.
It goes without saying that aircraft, ships, cars, etc. can also be used, and it can be used for general test equipment that requires diagnosis.

〔Effect of the invention〕

As explained above, the present invention connects a data diagnosis device to a conventional data processing device to effectively utilize design data during testing, and also functionalizes the manual test data diagnosis work. Diagnosis of satellite test data can be automated, which has the effect of saving labor and speeding up test work.

[Brief explanation of the drawing]

FIG. 1 is an overall system configuration diagram according to an embodiment of the present invention.
FIG. 2 is a software configuration diagram showing the software shown in the embodiment of this invention as a means, and FIG. 3 is a format 1-example of an FME person sheet created by the failure mode effect analysis chart creation means shown in FIG. Figure 4 explaining
FIG. 5 is a diagram illustrating the operation of the failure mode and effect analysis sheet creation means; FIG. 5 is a diagram illustrating the correspondence between the diagnostic rules created by the diagnostic rule creation means shown in FIG. 2 and the FMEA sheet; FIG. 6 is a diagram explaining the operation of the diagnostic rule creation means, FIG. 7 is a diagram explaining the knowledge database section shown in FIG. 2, and FIG. 8 is an overall system configuration diagram showing a conventional satellite testing device. It is. In the figure, (3) is a data processing device, (4) is a data diagnosis device, (J2) is a failure mode effect analysis sheet creation means, (1
3) is a diagnostic rule creation means, (14) is a diagnostic means, (1
5) is a knowledge database section, and (16) is an inference function section. Note that the same reference numerals in each figure indicate the same or corresponding parts.
Agent Masuo Oiwa Figure 1 3: Tee 1 Cutting device 4; De-Kaku diagnosis Figure 6 Figure 7 731 ■■ (Bottom part) ■ Hole No. 1, Power''; :2001f
<lma■ Book 1 - Wheel N0.1 Fat power 3250 I-) 1 (Magazine collapse leg Wheel N0.11 failure'*) There is. The problem is that the input level is low.

Claims (1)

[Claims] a data processing device for processing/editing test data of a device under test; a data diagnosis device connected to the data processing device for diagnosing the test data; and a test data processing means incorporated as software in the data processing device. A test device comprising: a failure mode effect analysis sheet creation means, a diagnostic rule creation means, and a diagnosis means, which are incorporated as software in the data diagnosis device.