CN106547248A - A kind of strategy-generating method for control system under open environment - Google Patents
A kind of strategy-generating method for control system under open environment Download PDFInfo
- Publication number
- CN106547248A CN106547248A CN201610880510.1A CN201610880510A CN106547248A CN 106547248 A CN106547248 A CN 106547248A CN 201610880510 A CN201610880510 A CN 201610880510A CN 106547248 A CN106547248 A CN 106547248A
- Authority
- CN
- China
- Prior art keywords
- description
- strategy
- stage
- natural language
- control
- 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.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B19/00—Programme-control systems
- G05B19/02—Programme-control systems electric
- G05B19/04—Programme control other than numerical control, i.e. in sequence controllers or logic controllers
- G05B19/05—Programmable logic controllers, e.g. simulating logic interconnections of signals according to ladder diagrams or function charts
- G05B19/056—Programming the PLC
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B2219/00—Program-control systems
- G05B2219/10—Plc systems
- G05B2219/13—Plc programming
- G05B2219/13004—Programming the plc
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Automation & Control Theory (AREA)
- Machine Translation (AREA)
Abstract
The invention discloses a kind of strategy-generating method for control system under open environment, including four-stage:First, the natural language description stage for application or field, control strategy is described with natural language mode or Domain Specific Language, export the natural language control strategy expression of specification;2nd, intermediate language describes the stage and the expression of natural language control strategy is converted to description in the middle of appropriate;3rd, performing the description stage is processed to middle description, the policy depiction data that the output of generation is performed for, and is performed by special or general regulation engine, is completed the output of control;4th, the communication description stage to it is above-mentioned for perform policy depiction data process, produce for communication policy depiction data.The description that can be made full use of internet with environment of internet of things to support natural language strategy or domain language strategy of the invention, reduces the difficulty and cost of intelligentized control method on the whole, reducing the technical threshold of user, facilitates the popularization of Automated condtrol application.
Description
Technical field
The present invention relates to Internet of Things and control field, and in particular to a kind of strategy for control system under open environment is raw
Into method.
Background technology
With the continuous popularization of Internet of Things application, opening imformation is introduced into traditional control field and has obtained people increasingly
Many concerns.These opening imformations include:Value chain related information, the sensing of control institute foundation and feedback information and control plan
The information of omiting.Under open network environment, the wisdom degree that above- mentioned information resource can greatly improve control is associated, reduce control
System processed possesses and use cost.And it is this be associated in current closed control system in be difficult to complete.Therefore open environment
The structure of lower control system is just particularly important.Number of patent application is 201610140028.4, entitled " under a kind of open environment
The domestic invention of control system and its implementation " proposes the control system under a kind of open environment, and the control system includes n
Layer control structure (n >=2), the make of every layer of control structure are identical, all including communication interface modules, communication protocol processes mould
Block, control data processing module, policy data processing module.The control system is the relation control structure of a stratification, phase
Adjacent bed time is realized interconnecting by data interaction.In the control structure of each layer, the function of policy data processing modules implement includes
The generation of policy data, forwarding and perform.If good strategy generation mechanism is not supported, the control under above-mentioned open environment
The advantage of system cannot perform to best, complicated control strategy application and just be difficult to popularize, and also have influence on Internet of Things application pair
The penetration of all trades and professions and people's daily life.And in current control field, most of strategy generation mechanisms are for envelope
Close and partial control system, in such a case it is difficult to give full play to opening imformation effect in the controlling and its advantage.
The content of the invention
The strategy generating that the technical problem to be solved in the present invention is adopted in being aiming at above-mentioned closing and partial control system
The deficiency of mechanism, proposes a kind of strategy-generating method for control system under open environment.
To solve above-mentioned technical problem, a kind of strategy for control system under open environment is the technical scheme is that
Generation method.Its main thought is the progressively formalization by describing to control strategy, and associating policy is generated in an automated manner
Each stage, reduce field description generate and final application use difficulty.
The course of work of the method includes the following four stage:
1) natural language description stage, this stage are directed to application or field, with natural language mode or Domain Specific Language
Description control strategy, final output are the natural language control strategy expression of specification;
2) intermediate language describes the stage, and it is appropriate that the natural language control strategy expression of above-mentioned specification is converted to by this stage
Middle description, the species of middle description depend on the field of control application and the content for specifically describing;
3) the description stage is performed, this stage is processed to description in the middle of above-mentioned, the plan that the output of generation is performed for
Data are slightly described, is performed by special or general regulation engine, complete the output for controlling;
4) communication description stage, this stage are processed to the above-mentioned policy depiction data for performing, and are produced for leading to
The policy depiction data of letter.
In the aforementioned natural language description stage, comprising the concrete steps that for control strategy is described in natural language mode:
1. the original sample of natural language description is collected, by the accumulation of these samples combination algorithm, key is summed up
Set of words and grammatical ruless, following steps 2.2 are coordination with step 2.3, may be selected to perform;
2. the form of navigation, on the basis of the 1st step, can be passed through, with the new natural language of man-machine interaction mode typing
The control strategy of description;
3. on the basis of the 1st step, it is also possible to by natural language recognition and handling implement limited keyword set
Original description is converted to the natural language description of specification.
Between in the preamble, language describes the stage, and the expression of natural language control strategy is converted to the concrete steps of middle description
It is:
1. according to keyword set extracting concept;
2. the method for being combined based on the correlation rule and template of key word is extracting the relation between concept;
3. according to the concept and the relationship of the concepts for extracting, and the intermediate language for selecting, carry out intermediate language and describe plan
Generation slightly.
Between in the preamble, language describes the stage, and alternative intermediate language includes but is not limited to ontology description language
OWL, finite state machine description, Petri-net descriptions, the description of ladder diagram text.
The description that the aforementioned output for performing the description stage is used in the policy depiction data for performing is set with concrete application
The realization of standby hardware and its embedded program is relevant, i.e., each different types of equipment should have a set of different switching method,
With the description for obtaining performing.
Beneficial effects of the present invention:
1, can make full use of internet with environment of internet of things to support retouching for natural language strategy or domain language strategy
State, so as to reducing the difficulty and cost of intelligentized control method on the whole, reducing the technical threshold of user, facilitate Automated condtrol
Using popularization.
2, the progressively formalization process of control strategy can make domain expert formulate control strategy convenient for control system
Ins and outs without understanding control system;Solve domain expert to be difficult to the knowledge of oneself is introduced asking for control system directly
Topic.
3, the source of non-formalization policy depiction sample, the product of sample size under open environment, can be greatly expanded
Tire out the application for artificial intelligences such as natural language processings to create conditions, these applications are to the popularization of Internet of Things application and wisdom
Realize it is critical that.And this it is this be accumulated in closed system in be difficult to realize.
4, the efficiency for generating improves in the automatization that implementation strategy is generated, and can make to be generated through network strategy transmission
To required strategy execution position.Solve description and the connection problem of realization of sophisticated strategies in Based Intelligent Control.
5, the present invention can be from the natural language description of strategy, and being formed by constantly formal process may finally
The policing rule set of execution.This contributes to for the description that domain knowledge is supported being automatically converted to the control strategy that can be performed
Description.
Description of the drawings
Fig. 1 is the schematic diagram of the strategy-generating method of the present invention.
Wherein:A represents the interface of man-machine interaction mode;B represents natural language recognition and the interface for processing;C is represented with
Between language come as control strategy description input;D represents the output in natural language description stage;E represents middle description;F tables
Show the policy depiction data for performing.
Specific embodiment
Below in conjunction with the accompanying drawings with one embodiment to the present invention be embodied as be described in further detail.
Embodiment is related to intelligent domestic system, illustrates the strategy-generating method of control system under open environment as example.
This strategy-generating method that the present invention is provided can lead to from the natural language of strategy or domain language description
Cross continuous formalization and form the policing rule set that may finally be performed.Final strategy can be by general or special purpose rule
Engine is performed.Whole generating process as shown in Figure 1, is divided into following four-stage:
1) the natural language description stage:The stage is directed to application or field, and domain expert or user can be with natures
Language or Domain Specific Language mode describe control strategy.Final output is the natural language control strategy expression of specification.It is raw
There are two kinds into mode:A kind of is the man-machine interaction by interface A, by way of navigation, continuous performance specification process;It is another kind of
It is the natural language for being converted to specification by the natural language recognition of interface B and process to the original description of limited keyword set
Description.In order that natural language is processed it is achieved that the keyword set that uses of description and grammatical ruless must be any limitation as.
The scope and grammatical ruless of set of keywords can determine according to the natural language processing process of varying level that the present invention does not do
Limit, this also embodies extensive adaptability of the invention.When natural language processing level is higher, field set of keywords credit union is big
Some, territory and intersection can also be larger, and rule can be with more wide in range;When natural language processing level is relatively low, field
Scope and intersection can be smaller, and field set of keywords credit union is smaller, and rule can be stricter.Under open environment, this stage
A vital task be to collect the original sample of natural language description, by the accumulation of these samples and with reference to appropriate calculation
Method so that the process of natural language policy depiction constantly can be evolved.
2) intermediate language describes the stage:The output D in natural language description stage is converted to appropriate centre by this stage
Description E, the species of middle description depend on the field of control application and the content for specifically describing.Generate the mesh of interstage description
Be natural language description and rule description between set up a bridge, can describe according to intermediate language, automatically generate control
System strategy.At the same time, in some cases, can be without natural language description directly with intermediate language come as control strategy
The input of description, as shown at c.Alternative intermediate language includes:Ontology description language OWL, finite state machine are described (such as
SDL, state-transition matrix etc.), Petri-net description, ladder diagram text description etc..Due to the output in natural language description stage
It is normalized, so can be processed using rule and method during intermediate language is generated.During process, place is needed
The extraction of reason concept and the relationship of the concepts.Concept is mainly according to keyword set being extracted, and the acquisition of the relationship of the concepts,
Can be extracted based on the method that the correlation rule of key word and template combine.
3) perform the description stage:The stage is to being input into E process, the policy depiction that the output F of generation is performed for
Data.Strategy is finally performed by special or general regulation engine, completes the output for controlling.For from principle, this stage
Actually a series of " If ... then ... " sentences of description.In addition, the output for performing the description stage is also used for depositing for strategy
Storage and strategy are also the storage in the form of performing the output in description stage.Perform the description in description stage and setting for concrete application
The realization of standby hardware and its embedded program is relevant, that is to say, that each different types of equipment should have a set of different turns
Method is changed, with the description for obtaining performing.For example, to Rete regulation engines, it is necessary to provide the description of regular collection, Cai Nengbao
Card performs the probability and correctness of description.
4) the communication description stage:The stage is used for the policy depiction number of communication to being input into F process, the output of generation
According to.Under open environment, the generator of strategy and user are probably different main bodys, are distributed in different geographical locations,
For example, the generator of the control strategy of intelligent domestic system can be manufacturer, and the user of strategy is user.In order that using
These strategies, can be by Network Capture, for that purpose it is necessary to consider the problem of ipsec communication.The communication description stage is retouched in execution
Increase appropriate expense on the basis of stating stage output, form the data structure suitable for communication transmission process.The stage retouches
It is a kind of data abstraction to state.It is described in different execution on form and transmission and unites, and is entered using various communication protocols
Row transmission, and it is transparent transmission for transmitting-receiving two-end, do not interfere with its understanding.
The description in aforementioned four stage can be distributed in the same control hierarchy of control method under open environment, so as to phase
Calling with situation, it is also possible to be distributed in different control hierarchys.In the natural language description stage, the generation of output has at one
Realize in having the control hierarchy of strategy generating/forwarding capability;Described in intermediate language the stage, the generation of output can be same
Or another has realization in the level of strategy generating/forwarding capability;The description stage is being performed, the generation of output can be in independence
Or with other stages share the level with strategy generating/forwarding capability in realize;In communication description stage, the life of output
Cheng Yi it is independent or share with other stages, realize in level undertaking tactful forwarding capability.
In specific application, the aforementioned four stage can be accepted or rejected according to different situations, for example, field description
In the case of highly formal, intermediate language can be directly entered and describe the stage, without through natural language description rank
Section;In the case where open network communication is not related to, the execution description stage can be directly entered without the communication description stage.With
This is further divided into different sub-phase process simultaneously inside above-mentioned stage, for example, the natural language description stage can be divided into people
Machine interacts stage and description generation phase.
This strategy-generating method also has following features:
1. the generation of natural language description stage output is real in having the control hierarchy of strategy generating/forwarding capability at one
It is existing;
2. intermediate language describe the stage output generation can it is same or another there is strategy generating/forwarding capability
Level in realize;
3. the generation of communication description stage output is independent at one or shares with other stages, undertakes forwarding capability
Realize in level;
4. perform the description stage output generate can it is independent or with other stages share with strategy generating/turn
Realize in sending out the level of function.
5. the execution of strategy is realized in the end Hierarchy nodes with variable policy enforcement function or intermediate level node.
6. it is be described can be stored in the controlling database of a certain level, so as to calling for same case.
Further illustrate the strategy of control system under open environment below by taking the signal light control in intelligent domestic system as an example
Generation system.
The first step:The natural language description stage.The smart home system of natural language description is constantly collected by open environment
Signal light control corresponding strategies in system;And during accumulation, carry out the extraction of key word and the setting of grammatical ruless.
This is the premise that correct parsing is carried out to the strategy that natural language mode is described.Strategy is exemplified below:
The signal light control strategy one that natural language mode is described:The instruction of " turning on light " is heard, is turned on light;Hear the finger of " turning off the light "
Order, turns off the light.
The signal light control strategy two that natural language mode is described:In the case that infrared induction someone and light intensity are less than 5,
Turn on light;In the case that infrared induction someone and light intensity are more than 5, turn off the light;In the case of infrared induction nobody, turn off the light.
After being analyzed to the above-mentioned course of work, the key word for extracting includes:Turn on light, turn off the light, people, he hears, instructs,
Have, do not have, situation, less than, more than, and, light intensity, infrared induction.
Second step:Intermediate language describes the stage.The angle developed from system design in this example, carries out systems organization, utilizes
Ontology Language is described in order to the description of natural language is converted to system object as intermediate language.The system object set
It is one and defines the definition such as strict data acquisition system, the incidence relation with attribute and and object between.Ontology Language description
Policy information can be stored in data base, be easy to the download and execution of strategy.The key word that the first step is summarized is analyzed
Afterwards, summing up needs the object for describing to be controlled device, on-off control event, heat transfer agent respectively.Wherein controlled device includes
Lamp, control event include open and close;Heat transfer agent includes:Voice messaging, light intensity, infrared induction value.After the completion of ontological analysis,
Open-Source Tools Jena can be utilized to complete the structure of body and the policy instance description based on body.
During second step, except in addition to key word conversion cost body language, also wanting the legitimacy (example of inspection policy
Such as whether have conflict, whether because of ambiguity or limit part that cannot not enough change etc.), such as find it is illegal describe, then need
Changed by second step or the corresponding information of supplement.
3rd step:Perform language and describe the stage.This step is converted to executable policy depiction data ontology describing.Example
Such as, if strategy execution system is PLC, ontology describing can just be converted into the structuring for using in PLC
The program that text language is write, can thus directly drive the work of strategy execution mechanism.
4th step:The communication description stage.This step is processed to the executable policy depiction data of above-mentioned 3rd step,
The policy depiction data for communication are produced, strategy execution part is then forwarded to, is reached the effect of control.In this step,
To be adapted to according to different communication protocol, it is ensured that the different types of network such as netting by IP network, Zigbee can be can
The policy depiction data is activation of execution is gone out.
Claims (5)
1. a kind of strategy-generating method for control system under open environment, it is characterised in that methods described includes following four
The individual stage:
1) the natural language description stage, for concrete application field, control strategy, final output are described in natural language mode
It is the natural language control strategy expression of specification;
2) intermediate language describes the stage, and the natural language control strategy expression of above-mentioned specification is converted to middle description, and centre is retouched
The species stated depends on the field of control application and the content for specifically describing;
3) the description stage is performed, description in the middle of above-mentioned is processed, the policy depiction data that the output of generation is performed for,
Performed by special or general regulation engine, complete the output for controlling;
4) the communication description stage, the above-mentioned policy depiction data for performing are processed, is produced and is retouched for the strategy for communicating
State data.
2. a kind of strategy-generating method for control system under open environment according to claim 1, it is characterised in that
Stage 1) in control strategy is described in natural language mode concrete steps include:
2.1. the original sample of natural language description is collected, by the accumulation of these samples combination algorithm, key word is summed up
Set and grammatical ruless, following steps 2.2 are coordination with step 2.3, may be selected to perform;
2.2. the form of navigation on the basis of 2.1 steps, can be passed through, is retouched with the natural language that man-machine interaction mode typing is new
The control strategy stated;
2.3. on the basis of 2.1 steps, it is also possible to by natural language recognition and handling implement the original of limited keyword set
Begin to describe the natural language description for being converted to specification.
3. a kind of strategy-generating method for control system under open environment according to claim 1, it is characterised in that
Stage 2) in the expression of natural language control strategy is converted to into comprising the concrete steps that for middle description:
3.1. according to keyword set extracting concept;
3.2. the method for being combined based on the correlation rule and template of key word is extracting the relation between concept;
3.3. according to the concept and the relationship of the concepts for extracting, and the intermediate language for selecting, carry out intermediate language Descriptive strategies
Generation.
4. a kind of strategy-generating method for control system under open environment according to claim 1, it is characterised in that
Stage 2) in alternative intermediate language include ontology description language OWL, finite state machine description, Petri-net description, ladder
Shape picture and text this descriptions.
5. a kind of strategy-generating method for control system under open environment according to claim 1, it is characterised in that
The device hardware of description and concrete application that the output for performing the description stage is used in the policy depiction data for performing and
The realization of its embedded program is relevant, i.e., each different types of equipment should have a set of different switching method, so that obtain can
With the description for performing.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610880510.1A CN106547248A (en) | 2016-10-09 | 2016-10-09 | A kind of strategy-generating method for control system under open environment |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610880510.1A CN106547248A (en) | 2016-10-09 | 2016-10-09 | A kind of strategy-generating method for control system under open environment |
Publications (1)
Publication Number | Publication Date |
---|---|
CN106547248A true CN106547248A (en) | 2017-03-29 |
Family
ID=58368543
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610880510.1A Pending CN106547248A (en) | 2016-10-09 | 2016-10-09 | A kind of strategy-generating method for control system under open environment |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106547248A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111104118A (en) * | 2018-10-29 | 2020-05-05 | 百度在线网络技术(北京)有限公司 | AIML-based natural language instruction execution method and system |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101710343A (en) * | 2009-12-11 | 2010-05-19 | 北京中机科海科技发展有限公司 | Body automatic build system and method based on text mining |
US20130151238A1 (en) * | 2011-12-12 | 2013-06-13 | International Business Machines Corporation | Generation of Natural Language Processing Model for an Information Domain |
CN103327080A (en) * | 2013-06-04 | 2013-09-25 | 北京大学 | Method and device for establishing control system of internet of things |
-
2016
- 2016-10-09 CN CN201610880510.1A patent/CN106547248A/en active Pending
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101710343A (en) * | 2009-12-11 | 2010-05-19 | 北京中机科海科技发展有限公司 | Body automatic build system and method based on text mining |
US20130151238A1 (en) * | 2011-12-12 | 2013-06-13 | International Business Machines Corporation | Generation of Natural Language Processing Model for an Information Domain |
CN103327080A (en) * | 2013-06-04 | 2013-09-25 | 北京大学 | Method and device for establishing control system of internet of things |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111104118A (en) * | 2018-10-29 | 2020-05-05 | 百度在线网络技术(北京)有限公司 | AIML-based natural language instruction execution method and system |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN111787090B (en) | Intelligent treatment platform based on block chain technology | |
CN111538842B (en) | Intelligent sensing and predicting method and device for network space situation and computer equipment | |
CN108446286A (en) | A kind of generation method, device and the server of the answer of natural language question sentence | |
CN103440287B (en) | A kind of Web question and answer searching system based on product information structure | |
Shi | [Retracted] Knowledge Graph Question and Answer System for Mechanical Intelligent Manufacturing Based on Deep Learning | |
CN106777222B (en) | Security equipment threat information sharing method based on lightweight domain ontology | |
CN105159148B (en) | Robot instruction processing method and device | |
CN113705218B (en) | Event element gridding extraction method based on character embedding, storage medium and electronic device | |
CN102685242A (en) | Method for establishing relevance relation among services under opened network | |
Myroshnichenko et al. | Mapping ER schemas to OWL ontologies | |
CN116776895A (en) | Knowledge-guided large language model query clarification method and system for API recommendation | |
CN106844934B (en) | Smart city planning and designing expert system and smart city planning and designing method | |
CN106547248A (en) | A kind of strategy-generating method for control system under open environment | |
CN107180024A (en) | A kind of multi-source heterogeneous data entity recognition methods of center connected subgraph and system | |
Kim et al. | General labelled data generator framework for network machine learning | |
CN102930030A (en) | Ontology-based intelligent semantic document indexing reasoning system | |
CN113220973B (en) | Public opinion verification method based on knowledge reasoning technology | |
CN110019768A (en) | Generate the method and device of text snippet | |
Li et al. | Research on optimization of knowledge graph construction flow chart | |
CN105279145A (en) | Semantic engineering system of computer | |
CN104572868A (en) | Method and device for information matching based on questioning and answering system | |
CN113742558A (en) | Query method, system, equipment and medium compatible and concurrent with multiple databases | |
Sixin et al. | A New Governance Architecture for Government Information Resources Based on Big Data Ecological Environment in China | |
CN103514217B (en) | Method and system for processing associated prompts of retrieval condition of retrieval application | |
CN117951314B (en) | Scenario generation decision method integrating knowledge graph and large language generation model |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170329 |
|
RJ01 | Rejection of invention patent application after publication |