JPH04135201A - Inferring method for expert system - Google Patents

Inferring method for expert system

Info

Publication number
JPH04135201A
JPH04135201A JP2258710A JP25871090A JPH04135201A JP H04135201 A JPH04135201 A JP H04135201A JP 2258710 A JP2258710 A JP 2258710A JP 25871090 A JP25871090 A JP 25871090A JP H04135201 A JPH04135201 A JP H04135201A
Authority
JP
Japan
Prior art keywords
knowledge
knowledge base
knowledge bases
inference
bases
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
JP2258710A
Other languages
Japanese (ja)
Other versions
JP2761090B2 (en
Inventor
Hideo Kuroda
英夫 黒田
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Mitsubishi Heavy Industries Ltd
Original Assignee
Mitsubishi Heavy Industries Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Mitsubishi Heavy Industries Ltd filed Critical Mitsubishi Heavy Industries Ltd
Priority to JP2258710A priority Critical patent/JP2761090B2/en
Publication of JPH04135201A publication Critical patent/JPH04135201A/en
Application granted granted Critical
Publication of JP2761090B2 publication Critical patent/JP2761090B2/en
Anticipated expiration legal-status Critical
Expired - Fee Related legal-status Critical Current

Links

Classifications

    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B29—WORKING OF PLASTICS; WORKING OF SUBSTANCES IN A PLASTIC STATE IN GENERAL
    • B29C—SHAPING OR JOINING OF PLASTICS; SHAPING OF MATERIAL IN A PLASTIC STATE, NOT OTHERWISE PROVIDED FOR; AFTER-TREATMENT OF THE SHAPED PRODUCTS, e.g. REPAIRING
    • B29C45/00—Injection moulding, i.e. forcing the required volume of moulding material through a nozzle into a closed mould; Apparatus therefor
    • B29C45/17—Component parts, details or accessories; Auxiliary operations
    • B29C45/76—Measuring, controlling or regulating
    • B29C45/766—Measuring, controlling or regulating the setting or resetting of moulding conditions, e.g. before starting a cycle

Landscapes

  • Engineering & Computer Science (AREA)
  • Manufacturing & Machinery (AREA)
  • Mechanical Engineering (AREA)
  • Injection Moulding Of Plastics Or The Like (AREA)
  • Devices For Executing Special Programs (AREA)
  • Feedback Control In General (AREA)

Abstract

PURPOSE:To output the counterplans to plural molding defects which occur at one time in consideration of each defect by applying the weight to plural knowledge bases, multiplying the variable value of the inferring result of each knowledge base by the weight, and outputting the weighted mean value. CONSTITUTION:An operator performs the necessary input operations through a keyboard 2 by reference to the display of a CRT 1. When the inferring results of >=2 knowledge bases are compounded and an answer is required among those knowledge bases KB1, KB2... of an entire knowledge base 7, the weight is applied to the corresponding knowledge base and the weighted mean is secured among the inferring results of knowledge bases. As a result, a compound inferring result is obtained with combination of plural knowledge bases. Thus it is possible obtain the answers of counterplans as the compound inferring result of knowledge bases in consideration of each defect against plural molding defects which occur at one time.

Description

【発明の詳細な説明】 （産業上の利用分野） 本発明は、射出成形エキスパートシステムその他各種の
エキスパートシステムにおいて、複数の知識ベースの推
論結果を合成した答を求める場合に適用できるエキスパ
ートシステムの推論方法に関するものである。DETAILED DESCRIPTION OF THE INVENTION (Field of Industrial Application) The present invention is an expert system inference that can be applied to an injection molding expert system and other various expert systems to obtain an answer by combining the inference results of multiple knowledge bases. It is about the method.

（従来の技術） 従来より、知識工学の発展によりコンピューターを用い
た種々のエキスパートシステムが開発、実用化されてい
る。ここで言うエキスパートシステムとは、知識ベース
と推論手続きを用いる、二とにより、特定の問題領域の
専門家と同等の問題解決能力を持つコンビ−’ｉミータ
−システムある。現在までのエキスパートシステムとし
ては、医療診断、故障診断、分子構造推定その他多数の
ものか開発されてきている。(Prior Art) With the development of knowledge engineering, various expert systems using computers have been developed and put into practical use. The expert system referred to herein is a combination system that uses a knowledge base and reasoning procedures, and has problem-solving abilities equivalent to those of experts in a particular problem area. Expert systems to date have been developed for medical diagnosis, failure diagnosis, molecular structure estimation, and many others.

本発明を含む一般的なエキスパートシステｊ７の概略構
成を第１図に示す。同図において、１はＣＲＴ　（Ｃａ
ｔｈｏｄｅ　Ｒａｙ　Ｔｕｂｅて所謂ブラウン管）、２
はキーボード、３は対象物の状態などを検出する各種セ
ンサー　４はコンピューター本体と前記ＣＲＴ１、キー
ボード２、センサー３とを繋ぐユーザーインターフェイ
スである。また５は制御システム、６は推論エンジン、
７は知識ベース（全体）で、これらはコンピューター本
体に組み込まれている。知識ベース（全体）７は各知識
ベースＫＢ、、ＫＢ２．−・の集合である。なお、１１
〜１７はそれぞれの要素を結ぶ信号線である。FIG. 1 shows a schematic configuration of a general expert system j7 including the present invention. In the same figure, 1 is a CRT (Ca
thode Ray Tube (so-called cathode ray tube), 2
3 is a keyboard, 3 is various sensors for detecting the state of objects, etc. 4 is a user interface that connects the computer main body with the CRT 1, keyboard 2, and sensor 3. 5 is a control system, 6 is an inference engine,
7 is the knowledge base (overall), which is built into the computer itself. The knowledge base (whole) 7 includes each knowledge base KB, KB2 . It is a set of -・. In addition, 11
17 are signal lines connecting each element.

本エキスパートシステムの推論実行は、前記各知識ベー
スＫ　Ｂ　ｌ＋　Ｋ　Ｂ　２．−の何れかを呼び出し、
推論エンジン６により行なう。この推論の結果として、
取り上げた問題に対する答が得られる。The inference execution of this expert system is based on each of the knowledge bases K B l+ K B 2. - call one of
This is done by the inference engine 6. As a result of this reasoning,
You will get answers to the questions raised.

一方プラスチックの射出成形不良対策のエキスパートシ
ステムては、成形不良の種類毎の対策知識を前記各知識
ベースＫ　Ｂ　、、　Ｋ　Ｂ　２．−　　として用意し
、発生した成形不良に対応する知識ベースを選択して推
論する。従って単一の成形不良に対しては問題ないか、
複数の成形不良か同時に発生する場合は、従来次の何れ
かの方法を取らざるを得なかった。On the other hand, an expert system for countermeasures against defects in plastic injection molding uses the knowledge bases K B , K B 2. - Select and infer the knowledge base corresponding to the molding defect that has occurred. Therefore, is there no problem with a single molding defect?
Conventionally, when multiple molding defects occur simultaneously, one of the following methods has to be taken.

（ａ）　　どれか最も重大な成形不良のみを取り上げる
。(a) Pick up only the most serious molding defects.

（ｂ）　　複数の成形不良の組合せ毎に対策の知識ベー
スを作成し直す。(b) Recreate the knowledge base for countermeasures for each combination of multiple molding defects.

しかるに（ａ）の方法では、成形不良か複数発生してい
るのに、１個の成形不良しか解決できなかった。また（
ｂ）の方法では、複数の成形不良の組合せ数が多すぎて
対策の知識ベースの作成に手間がかかりすぎるという欠
点があった。However, in method (a), only one molding defect could be resolved even though multiple molding defects had occurred. Also(
The method b) has the disadvantage that the number of combinations of multiple molding defects is too large and it takes too much effort to create a knowledge base for countermeasures.

（発明が解決しようとする課題） 上述のように、従来のエキスパートシステムでは、個別
の知識ベースを推論して個別の答を出すことしかできな
かったので、射出成形不良対策のエキスパートシステム
で同時発生の複数の成形不良に対して対策を出したい場
合に、複数の成形不良には全く対応てきないか、或いは
複数の成形不良の組合せ多数について膨大な知識ベース
を作成しなければならないという問題点かあった。(Problem to be solved by the invention) As mentioned above, conventional expert systems could only infer individual knowledge bases and provide individual answers. If you want to take countermeasures against multiple molding defects, the problem is that you cannot deal with multiple molding defects at all, or you have to create a huge knowledge base for many combinations of multiple molding defects. there were.

（課題を解決するだめの手段） このため本発明は、複数の知識ベースの推論結果を合成
した答を求めるエキスパートシステムにおいて、該当の
複数の知識ベースに重み付けをすると共に、各知識ベー
ス毎の推論結果の各変数値に前記重みを掛け、これらの
結果に基づいて重み付き平均した値を出力するもので、
これを課題解決のための手段とするものである。(Means for Solving the Problem) Therefore, in an expert system that obtains an answer by combining the inference results of a plurality of knowledge bases, the present invention weights the plurality of knowledge bases and makes inferences for each knowledge base. It multiplies each resulting variable value by the weight and outputs a weighted average value based on these results.
This is a means to solve problems.

（作用） 該当する各知識ベースに重み付けし、各知識ベースの推
論結果に対して重み付き平均を取るか、又は重み合計の
最大のものを答とすることにより、従来の知識ベースの
ままで、複数の知識ベースを組合せた複合推論結果を得
ることができる。従って、射出成形不良対策のエキスパ
ートシステムで、同時発生する複数の成形不良に対して
、各不良毎の知識ベースの複合推論結果として各不良を
考慮した対策を答に得られる。(Function) By weighting each applicable knowledge base and taking the weighted average of the inference results of each knowledge base, or by taking the maximum weight total as the answer, the conventional knowledge base can be used as is. Composite inference results can be obtained by combining multiple knowledge bases. Therefore, in an expert system for dealing with injection molding defects, countermeasures can be obtained that take each defect into consideration as a composite inference result based on the knowledge base for each defect in response to multiple molding defects that occur simultaneously.

また各知識ベースの重み付けの値は状況に応じて変える
ことかできるので、その時々で重要な知識ベースと重要
でない知識ベースを差別して扱える。Furthermore, since the weighting value of each knowledge base can be changed depending on the situation, important knowledge bases and unimportant knowledge bases can be treated differently at any given time.

（実施例） 以下、図面に基づいて本発明の詳細な説明すると、第１
図は従来技術と共通で、−船釣なエキスパートシステム
の概略構成を示し、従来の技術のところで説明したもの
と同じである。(Example) Hereinafter, the present invention will be described in detail based on the drawings.
The figure is common to the prior art, and shows a schematic configuration of a fishing expert system, which is the same as that described in the prior art section.

第２図は知識ベースの推論結果が各変数Ｘ　の一定値で
与えられる場合に、各変数Ｘ　について各知識ベースの
重み付き平均Ｘ　ＩＡを計算するフロー図を示す。また
第３図は知識ベースの推論結果が変数Ｘ１の値の範囲で
与えられる場合に、各変数Ｘ、ｌこついて各知識ベース
の重みの合計最大の区間Ｘ　１ＡＬ−Ｘ　ＩＡＬＩを計
算するフロー図を示す。FIG. 2 shows a flow diagram for calculating the weighted average X IA of each knowledge base for each variable X 2 when the inference result of the knowledge base is given by a constant value of each variable X 2 . Fig. 3 is a flowchart for calculating the maximum interval X1AL-XIALI of the sum of the weights of each knowledge base for each variable shows.

第１図において、操作者はＣＲＴｌの表示を見、キーボ
ード２から必要な入力を行なう。知識ベース（全体）７
の各知識ベースＫ　Ｂ　＋、　Ｋ　Ｂ　２゜・・−の内
、２つ以上の各知識ベースの推論結果を合成した答を求
める必要がある場合は、操作者か該当の各知識ベースに
その時の状況に応じた重みを与え、第２図又は第３図の
フローに従って計算する。In FIG. 1, the operator looks at the display on the CRTl and makes necessary inputs from the keyboard 2. Knowledge base (overall) 7
If it is necessary to obtain an answer by combining the inference results of two or more knowledge bases among the knowledge bases K B +, K B 2゜...-, the operator or each knowledge base must A weight is given according to the situation, and the calculation is performed according to the flow shown in FIG. 2 or 3.

次に第２図について説明すると、ｎ個の知識ベースＫＢ
、、ＫＢ２．−．ＫＢ、を使用し、推論結果１ｉｒｒｈ
個の変数ｘ　、、　ｘ　２．−、　ｘ、について各知識
ベース毎に値が得られるものとする。第２図のステップ
２１では、上述のｎ個の知識ベースとその各知識ベース
ＫＢ、に対する重みＷ、を定めている。次にステップ２
２とステップループ２３〜２５で、各知識べ・−スＫ　
Ｂ　、の推論を実行し、各推論結果として各変数Ｘ１の
値ｘｉ＋を得る。更にステップ２６で、各変数Ｘ１につ
いて重み付き平均Ｘ　＋　Ａを計算し、工程ステップで
それら重み付き平均Ｘ　ＩＡを答として出力する。Next, to explain Figure 2, n knowledge bases KB
,,KB2. −． Using KB, the inference result is 1irrh
variables x,, x2. −, x, values are obtained for each knowledge base. In step 21 of FIG. 2, the weight W for the above-mentioned n knowledge bases and each knowledge base KB is determined. Next step 2
2 and step loops 23 to 25, each knowledge base K
B, and obtain the value xi+ of each variable X1 as a result of each inference. Further, in step 26, a weighted average X + A is calculated for each variable X1, and in a process step, the weighted average X IA is outputted as an answer.

同様に第３図について説明すると、ステップ３１は第２
図のステップ２１と同じである。ステップ３２とステッ
プループ３３〜３５て、各知識ベースＫＢ、の推論を実
行し、各推論結果として各変数Ｘ、の値の範囲Ｘ　Ｉ　
Ｉ　Ｌ−Ｘ　＋　Ｉ　Ｉ＋を得る。次にステップ３６と
ステップループ３７〜４０に移り、ステップ３７て各変
数Ｘ、毎に上記変数値−１，限Ｘ　ｉ　Ｉ　ｕ、下限Ｘ
　Ｉ　Ｉ　ＬをＸ１軸上に取り区間分けする。それらの
区間を■、■、−・−とじ、変数値範囲ＸｚＬ〜Ｘ　＋
＋ｕに重みＷ、かあるとして、ステップ３８で各区間■
、■９−毎に各知識ベース毎の重みＷ、を合計し、重み
合計が最大の区間Ｘ　ＩＡＬ−Ｘ　ＩＡＩＩを各変数Ｘ
１毎に求める。結果をステップ４１で出力する。Similarly, explaining FIG. 3, step 31 is the second step.
This is the same as step 21 in the figure. In step 32 and step loops 33 to 35, inference is executed for each knowledge base KB, and the value range X I of each variable X is determined as each inference result.
Obtain I L−X + I I+. Next, the process moves to step 36 and step loops 37 to 40, and in step 37, for each variable X, the above variable value -1, limit X i I u, lower limit X
Take I I L on the X1 axis and divide it into sections. Bind those intervals to ■, ■, −・−, and set the variable value range XzL to X +
Assuming that +u has a weight W, each interval ■
,■Sum up the weights W for each knowledge base for every 9-, and set the interval X IAL-X IAII with the maximum weight total to each variable X
Find every 1. The results are output in step 41.

（発明の効果） 以上述べたように本発明によれば、複数の知識ベースの
推論結果を合成した答を得ることができるので、射出成
形不良対策エキスパートシステムで各成形不良毎に知識
ベースを構成し７た場合でも、同時発生した複数の成形
不良に対して各不良を考慮した対策を出力できる。従っ
て従来各成形不良毎の知識ベースでは、同時発生した複
数の成形不良に対して対策を得られなかった問題が解決
できる。(Effects of the Invention) As described above, according to the present invention, it is possible to obtain an answer by combining the inference results of multiple knowledge bases, so the injection molding defect countermeasure expert system configures a knowledge base for each molding defect. Even in the case of multiple molding defects that occur simultaneously, countermeasures can be output that take each defect into consideration. Therefore, the conventional knowledge base for each molding defect can solve the problem of not being able to take measures against multiple molding defects that occur simultaneously.

また各知識ベースに重み付けを行なうので、その時々の
状況（成形不良の状況なと）に応じて重要な知識ベース
と重要でない知識ベースを差別でき、的確な答を得るこ
とかできる。以上の効果は射出成形不良対策エキスパー
トシステムのみならず、複数の知識ベースの推論結果を
合成する必要かある場合は、一般のとのエキスバー）シ
ステムにもあてはまるものである。Furthermore, since each knowledge base is weighted, it is possible to distinguish between important knowledge bases and unimportant knowledge bases depending on the situation at the time (such as the situation of molding defects), making it possible to obtain accurate answers. The above effects apply not only to injection molding defect countermeasure expert systems, but also to general expert systems when it is necessary to synthesize inference results from multiple knowledge bases.

【図面の簡単な説明】[Brief explanation of the drawing]

第１図は本発明を含む一般的なエキスパートシステムの
概略構成図、第２図は本発明の実施例を示す知識ベース
の推論結果か各変数の一定値て与えられる場合の重み付
き平均計算フロー図、第３図は本発明の実施例を示す知
識ベースの推論結果か各変数の値の範囲で与えらねる場
合の重み合計最大区間計算フロー図である。 図の主要部分の説明 １−ＣＲＴ ２−キーボード ４　ユーザー−インターフェイス ５　制御システム ６−推論エンジン ７−知識ベース（全体） ＫＢ、−・−各知識ベース Ｗ、・−・−各知識ベースの重み Ｘｌ・−推論結果の各変数 Ｘ　ＩＡ−・−各変数の重み付き平均 Ｘ　ＩＡＬ”””Ｘ　ＩＡｔ＋’−・−各変数の重み合
計最大の区間晃１ ノ に８４Figure 1 is a schematic configuration diagram of a general expert system including the present invention, and Figure 2 is a weighted average calculation flow when knowledge-based inference results or constant values of each variable are given, showing an embodiment of the present invention. 3A and 3B are flowcharts for calculating the maximum interval of weight sum when the knowledge-based inference result cannot be given within the value range of each variable, showing an embodiment of the present invention. Description of main parts of the diagram 1-CRT 2-Keyboard 4 User-interface 5 Control system 6-Inference engine 7-Knowledge base (overall) KB, ---Each knowledge base W, ---Weight Xl of each knowledge base - Each variable of the inference result IA - - Weighted average of each variable

Claims (2)

【特許請求の範囲】[Claims] （１）複数の知識ベースの推論結果を合成した答を求め
るエキスパートシステムにおいて、該当の複数の知識ベ
ースに重み付けをすると共に、各知識ベース毎の推論結
果の各変数値に前記重みを掛け、これらの結果に基づい
て重み付き平均した値を出力することを特徴とするエキ
スパートシステムの推論方法。(1) In an expert system that obtains an answer by combining the inference results of multiple knowledge bases, the multiple knowledge bases are weighted, each variable value of the inference result for each knowledge base is multiplied by the weight, and these An expert system inference method characterized by outputting a weighted average value based on the results of. （２）複数の知識ベースの推論結果を合成した答を求め
るエキスパートシステムにおいて、該当の複数の知識ベ
ースに重みを付け、各知識ベース毎の推論結果の各変数
値区間の上下限で区切られる各区間について知識ベース
毎の重み合計を求め、前記最も大きい重み合計を有する
区間を出力することを特徴とするエキスパートシステム
の推論方法。(2) In an expert system that obtains an answer by combining the inference results of multiple knowledge bases, the multiple knowledge bases are weighted, and each of the inference results for each knowledge base is divided by the upper and lower limits of each variable value interval. An inference method for an expert system, characterized in that the total weight for each knowledge base is determined for an interval, and the interval having the largest total weight is output.
JP2258710A 1990-09-27 1990-09-27 Expert System Reasoning Method Expired - Fee Related JP2761090B2 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
JP2258710A JP2761090B2 (en) 1990-09-27 1990-09-27 Expert System Reasoning Method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
JP2258710A JP2761090B2 (en) 1990-09-27 1990-09-27 Expert System Reasoning Method

Publications (2)

Publication Number Publication Date
JPH04135201A true JPH04135201A (en) 1992-05-08
JP2761090B2 JP2761090B2 (en) 1998-06-04

Family

ID=17324021

Family Applications (1)

Application Number Title Priority Date Filing Date
JP2258710A Expired - Fee Related JP2761090B2 (en) 1990-09-27 1990-09-27 Expert System Reasoning Method

Country Status (1)

Country Link
JP (1) JP2761090B2 (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111783931A (en) * 2020-06-04 2020-10-16 深圳市酷开网络科技有限公司 Internet of things expert system implementation method and system based on hybrid reasoning
CN112036571A (en) * 2020-08-25 2020-12-04 深圳市酷开网络科技有限公司 Multi-field expert system joint reasoning method and system

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPS63279302A (en) * 1987-05-12 1988-11-16 Fujitsu Ltd Fuzzy inference system
JPH02176802A (en) * 1988-12-27 1990-07-10 Toshiba Corp Fuzzy controller

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPS63279302A (en) * 1987-05-12 1988-11-16 Fujitsu Ltd Fuzzy inference system
JPH02176802A (en) * 1988-12-27 1990-07-10 Toshiba Corp Fuzzy controller

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111783931A (en) * 2020-06-04 2020-10-16 深圳市酷开网络科技有限公司 Internet of things expert system implementation method and system based on hybrid reasoning
CN112036571A (en) * 2020-08-25 2020-12-04 深圳市酷开网络科技有限公司 Multi-field expert system joint reasoning method and system
CN112036571B (en) * 2020-08-25 2023-12-29 深圳市酷开网络科技股份有限公司 Multi-domain expert system joint reasoning method and system

Also Published As

Publication number Publication date
JP2761090B2 (en) 1998-06-04

Similar Documents

Publication Publication Date Title
US4730259A (en) Matrix controlled expert system producible from examples
US6456989B1 (en) Neuro-fuzzy-integrated data processing system
US5043915A (en) Knowledge-based system installed with associative knowledge and inference method
JPH04135201A (en) Inferring method for expert system
JP2006268813A (en) Learning control device, learning control method, and program
US5150367A (en) Composite range constraint propagation control
Lin et al. Integration of CMAC technique and weighted regression for efficient learning and output differentiability
JPH07234988A (en) Abnormality diagnostic device
JPH04149663A (en) Inverse model generating method and control system applying same generating method
JPH05197705A (en) Learning system for neural network
JPH0490076A (en) Method for extracting feature of medical image
JP2778309B2 (en) Knowledge-based computer system
JPH04291662A (en) Operation element constituted of hierarchical network
JPH0371303A (en) Fuzzy controller
JPH04233657A (en) Instructing method for regression neural network
JPH0628192A (en) Fuzzy inference device
JP3101742B2 (en) Minimum detector
JPH0358132A (en) Inference control system
JP2559878B2 (en) Fuzzy controller
Farrokhi et al. Application of Prony signal analysis to recurrent neural networks
JP2559880B2 (en) Fuzzy controller
KR100224407B1 (en) Substructuring method for dynamic structure modification
JPH0320864A (en) Information processor having vector instruction set for fuzzy calculation
JPH0391030A (en) Method and device for debugging for fuzzy inference
JPH02244331A (en) Fuzzy inference device

Legal Events

Date Code Title Description
LAPS Cancellation because of no payment of annual fees