Summary of the invention
The present invention has pushed the management of concrete dam construction and decision-making work to intelligent pattern from traditional mode, no
Only be conducive to improving work efficiency, be more beneficial for overcoming the mistake contingent with decision-making or improper of labor management in traditional mode
Part, is effectively improved rationalization level and scientific degree, it is ensured that the management level of concrete dam construction, improves
Construction quality, accelerate construction speed, improve the comprehensive benefit of relevant industries and department, for playing concrete dam dam type
Comprehensive advantage will play a significant role.
The technical solution used in the present invention:
A kind of concrete dam construction intelligent decision system of KBE, this system include engineering information storehouse,
Construction rule base, intelligent inference system and functional module system;Engineering information storehouse includes arrangement and method for construction information, job site information
With the data of three aspects of dam body model information, the basic data support providing comprehensive is run in engineering information storehouse for system;Construction
Rule base is then responsible for the expert in the Its Relevant Technology Standards of storage and management concrete dam construction, technology requirement and industry
Experience and the rule knowledge of other concrete dam construction;The thinking process of the intelligent inference system simulation mankind, with rule of constructing
Then the rule in storehouse is reasoning foundation, and according to the related data in engineering information storehouse, based on data, reasoning obtains being correlated with
Conclusion or constructing operation;Functional module system includes position in storehouse orchestration module, machinery scheduler module, plan-validation module, progress mould
Intend module, schedule module of getting the raw materials ready, grouting schedule module and water flowing scheme module.
Above-mentioned engineering information storehouse specifically includes construction environment characteristic parameter, dam structure characteristic parameter, construction machinery spy
Levy parameter, construction technical requirement parameter, material parameter, arrangement and method for construction parameter and engineering construction parameter.
Above-mentioned construction environment characteristic parameter includes rainfall, snowfall, water temperature and flow;Dam structure characteristic parameter bag
Include monolith structural parameters, position in storehouse spatial parameter, hollow spaces parameter and concrete performance parameter;Construction machinery characteristic parameter includes
The space orientation of concrete warehousing construction machinery, service behaviour, operation characteristic and working range;Construction technical requirement parameter includes
The dam concrete pouring construction discrepancy in elevation limits, the intermittent time limits and temperature-controlled parameter;Material parameter includes dam concrete
The production of match ratio, sandstone and storage system and concrete store system;Arrangement and method for construction parameter include concreting schedule requirement,
Mechanization and special technical requirement;Engineering construction parameter includes on-the-spot the pouring position in storehouse, fill of concrete dam construction
Slurry/water flowing position, respectively put in storage machinery current location.
Above-mentioned construction rule base specifically includes the requirement of dam concrete mix technology, concrete transportation technology requires, dam
Body concrete pours rule, concrete water flowing cooling rule and dam body joint grouting rule.
Above-mentioned dam concrete mix technology requires to include concrete type, refrigeration requires and duration of mixing;Concrete
Transportation technology requires to include the free-electron model height in transportation, transhipment number of times and haulage time dam concrete and pours rule
Including discrepancy in elevation restriction, intermittent time and curing time in concreting process;Concrete water flowing cooling rule includes dam body
The beginning and ending time of concrete water flowing, trigger condition, water temperature and flow;Dam body joint grouting rule includes that dam concrete coordinates
The production of ratio, sandstone and storage system and concrete store system;Arrangement and method for construction parameter include dam body joint grouting time started,
Trigger condition and grouting region.
The decision method of the concrete dam construction intelligent decision system of a kind of KBE, it includes walking as follows
Rapid:
(1) conventionally form represents the premise of knowledge and the extraction of conclusion,
In current concrete dam construction management with decision making process, whether a certain position in storehouse concrete can pour needs
Meet the restriction of before and after's discrepancy in elevation, the left and right discrepancy in elevation limits, the interlayer intermittent time limits, meteorological condition limits, adjacent position in storehouse is constructed about
Restraint these to limit and constraints;When all of precondition meets, this position in storehouse can carry out pouring construction;From knowledge work
The production rule angle of journey understands, with constraints, these limit can regard whether a certain position in storehouse can pour this rule as
Precondition, " can pour " is then corresponding to the conclusion of this rule;
In the step for knowledge is extracted, it is whether the concrete position in storehouse represented with conventionally form can pour this
One judges that work is converted into the premise part of rule, conclusion part, reliability form;Before whether concrete position in storehouse can pour
Before and after's discrepancy in elevation, the left and right discrepancy in elevation, intermittent time, meteorological condition, 5 aspects of adjacent construction are put forward;Conclusion is for pouring or can not
To pour, for 0-1 variable;According to concrete dam construction specification, this rule is completely reliable;
(2) rule premise, conclusion, the quantization of reliability, classification,
On the basis of (one) step work, carry out quantifying, classifying or classification by the premise of rule, conclusion, reliability;
The discrepancy in elevation front and back: represent with absolute physical value, i.e. represents with the actual discrepancy in elevation of front and back's position in storehouse, passes through Formal Languages
Be expressed as Height_Diff-FB_Bin (x, y);The discrepancy in elevation before and after i.e. position in storehouse x and position in storehouse y, position in storehouse x higher than position in storehouse y be on the occasion of,
Otherwise it is then negative value;
The left and right discrepancy in elevation: represent with absolute physical value, i.e. represents with the actual discrepancy in elevation of left and right position in storehouse, passes through Formal Languages
Be expressed as Height_Diff-LR_Bin (x, y);The i.e. left and right discrepancy in elevation of position in storehouse x and position in storehouse y, position in storehouse x higher than position in storehouse y be on the occasion of,
Otherwise it is then negative value;
Intermittent time: represent with absolute physical value, both poured the deadline plus rule with the bottom position in storehouse of current position in storehouse
On the basis of the intermittent time that model requires or current engineering sets, it is judged that whether current time had arrived in this time period;Pass through
Formal Languages is represented by Pause_time (x), and its value is absolute calendar time, its form be " MM/dd/yyyy hh:
mm;ss”;
Meteorological condition: whether current position in storehouse can be poured by meteorological condition mainly considers from rainfall angle, can be right
Rainfall carries out classification, as heavy rain, heavy rain, light rain, without rain;Being represented by Rain (t) by Formal Languages, " t " i.e. herein
For drafting the time of moduling plan, the value of Rain (t) is according to the weather forecast parameter determination in data base;
Adjacent position in storehouse construction constraint: the construction constraint of adjacent position in storehouse refers to when certain position in storehouse casting concrete, adjacent thereto
Other positions in storehouse because draining, template erection and the cause of failures that puts in storage cannot carry out concrete pouring construction simultaneously;This is constrained to
0-1 variable bound, when there are adjacent position in storehouse just at casting concrete, this position in storehouse can not casting concrete, this binding occurrence is
0;Otherwise, all positions in storehouse adjacent thereto are not the most at casting concrete, then current position in storehouse can carry out concrete pouring construction, should
Binding occurrence is 1;It is represented by Adjoin_Bin (x) by Formal Languages;Which position in storehouse to be the adjacent storehouse of current position in storehouse as
Position, determines according to the three-dimensional relationship of dam body model parameter;
Conclusion: whether can pour the judgement of sets of conditions according to above current position in storehouse, conclusion is for pouring or can not
To pour, for 0-1 variable;By Formal Languages, FitPour-Bin (x) can be expressed as;
Reliability: according to concrete dam construction specification, this rule is completely reliable, i.e. reliability is 1:CF=1;
(3) formalization of production rule,
In the concrete dam construction intelligent decision system of KBE, before rule in (two) step
Carry, the quantization of conclusion, reliability, classification, and the statement of Formal Languages, " during concrete dam construction certain
Whether one position in storehouse concrete can pour under current state " judgement of this event can be expressed as following production rule
Then:
The precondition of this rule be simultaneously meet before and after the discrepancy in elevation, the left and right discrepancy in elevation, intermittent time, meteorological condition, adjacent storehouse
Position construction constraint, it may be assumed that P=Height_Diff-FB_Bin (x, y) ∧ Height_Diff-LR_Bin (x, y) ∧ Pause_time
(x)∧Rain(t)∧Adjoin_Bin(x);The conclusion of this rule is that current position in storehouse can be with casting concrete, it may be assumed that C=
FitPour-Bin(x);The credibility of rule is 1;So, current rule can be expressed as:
Height_Diff-FB_Bin(x,y)∧Height_Diff-LR_Bin(x,y)∧Pause_time(x)∧Rain
(t)∧Adjoin_Bin(x)→FitPour-Bin(x)(1)
(4) rule inspection,
The inspection of rule is the important step of production rule acquisition work, be the examination to abovementioned steps gained knowledge and
Check on;In this step, major tasks are that by reflecting that this rule is verified by the example of various condition of construction, look into
See that its conclusion exported is the most correct;For the output incorrect situation of conclusion, need check its reasoning process in detail and find
Reason;In rule inspection, it is accordingly required in particular to verify that the adaptability of this rule is with correct by some construction parameters in particular cases
Property;If after system test, do not find current rule existing problems, then this rule warehouse-in is preserved;If it find that exist
Problem, then need again to extract current rule and formalization, i.e. repeats step ()-step (three);
(5) again extract or formalization, for being found to have the rule of reasoning impropriety in checkout procedure, repeat step
Suddenly (one)-step (three), extract or formalization again to this rule;
(6) warehouse-in, for rule qualified after inspection, is deposited in rule base, with the form of relational database
Preserve.
The technique effect that the present invention obtains:
1, the concrete dam construction Intelligent Decision-making Method of the KBE that the present invention is set up and knowledge based
The concrete dam construction intelligent decision system of engineering, in the premise following concrete dam construction technical specification and requirement
Under, by the application of the methods such as the representation of knowledge of knowledge engineering, knowledge acquisition and knowledge reasoning, realized by intelligent inference system
Management in concrete dam construction and decision-making work.The method and system that the present invention sets up achieves concrete dam pouring and executes
Integrated, the automatization of work management and decision-making work, intellectuality, improve efficiency and the science of management and decision-making.
2, the concrete dam construction intelligent decision system of KBE that the present invention sets up has and pours position in storehouse
Layout, machinery scheduling of putting in storage, arrangement and method for construction are verified, are poured progress simulation, the concrete plan of getting the raw materials ready, grouting plan, water flowing
The function of seven aspects such as solution formulation, can meet the field management of concrete dam construction and the multinomial work of decision-making.And
Native system is prone to foundation, easy to use, maintenance is simple, highly versatile, the degree of reliability are high.
3, the present invention has pushed the management of concrete dam construction and decision-making work to intelligent pattern from traditional mode,
Not only contribute to improve work efficiency, be more beneficial for overcoming in traditional mode labor management and the contingent mistake of decision-making or not
Work as part, be effectively improved rationalization level and scientific degree, it is ensured that the management level of concrete dam construction, carry
High construction quality, accelerate construction speed, improve the comprehensive benefit of relevant industries and department, for playing concrete dam
The comprehensive advantage of type will play a significant role.Generally speaking, the present invention not only has good economic outlook, more has good
Social benefit.
Detailed description of the invention
Below in conjunction with the accompanying drawings embodiments of the present invention are described further.
See Fig. 1,3,5,6, the concrete dam construction intelligent decision system of a kind of KBE, this system
Including engineering information storehouse, construction rule base, intelligent inference system and functional module system;Engineering information storehouse includes that arrangement and method for construction is believed
Breath, job site information and the data of three aspects of dam body model information, the base providing comprehensive is run in engineering information storehouse for system
Plinth data are supported;Construction rule base is then responsible for the Its Relevant Technology Standards of storage and management concrete dam construction, technology requirement
And the expertise in industry and the rule knowledge of other concrete dam construction;The think of of the intelligent inference system simulation mankind
Dimension process, with the rule constructed in rule base for reasoning foundation, according to the related data in engineering information storehouse, with data as base
Plinth, reasoning obtains conclusion or the constructing operation being correlated with;Functional module system includes position in storehouse orchestration module, machinery scheduler module, side
Case authentication module, progress analog module, schedule module of getting the raw materials ready, grouting schedule module and water flowing scheme module.
Described engineering information storehouse specifically includes construction environment characteristic parameter, dam structure characteristic parameter, construction machinery spy
Levy parameter, construction technical requirement parameter, material parameter, arrangement and method for construction parameter and engineering construction parameter.Described construction ring
Border characteristic parameter includes rainfall, snowfall, water temperature and flow;Dam structure characteristic parameter includes monolith structural parameters, position in storehouse
Spatial parameter, hollow spaces parameter and concrete performance parameter;Construction machinery characteristic parameter includes concrete warehousing construction machinery
Space orientation, service behaviour, operation characteristic and working range;Construction technical requirement parameter includes dam concrete pouring construction
The discrepancy in elevation limits, the intermittent time limits and temperature-controlled parameter;Material parameter includes that dam concrete match ratio, sandstone produce and deposit
Storage system and concrete storage system;Arrangement and method for construction parameter includes concreting schedule requirement, mechanization and special technique
Requirement;Engineering construction parameter includes on-the-spot the pouring position in storehouse, be in the milk/water flowing position, respectively put in storage of concrete dam construction
Machinery current location.
Described construction rule base specifically includes the requirement of dam concrete mix technology, concrete transportation technology requires, dam
Body concrete pours rule, concrete water flowing cooling rule and dam body joint grouting rule.Described dam concrete mix skill
Art requires to include concrete type, refrigeration requires and duration of mixing;Concrete transportation technology requires to include oneself in transportation
The discrepancy in elevation limit that rule includes in concreting process is poured by falling height, transhipment number of times and haulage time dam concrete
System, intermittent time and curing time;Concrete water flowing cooling rule includes the beginning and ending time of dam concrete water flowing, triggers bar
Part, water temperature and flow;Dam body joint grouting rule includes the production of dam concrete match ratio, sandstone and storage system and concrete
Storage system;Arrangement and method for construction parameter includes the time started of dam body joint grouting, trigger condition and grouting region.
The major function of this system includes:
Pour position in storehouse orchestration module: concrete dam construction management, with decision making process, needs to water according to current dam body
Build situation, formulate continually each monolith, each bin level pour order, i.e. position in storehouse layout;And position in storehouse layout needs to meet work
Journey schedule requirement, image schedule requirement, and a series of associated construction process constraint;The concrete dam of KBE waters
Build construction intelligent decision system and can meet associated construction specification, dam according to the integrated information of current job site each side
Body concrete pouring technique requires, on the premise of progress and image schedule requirement, inference system can formulate each that not yet pour
Pouring sequentially of position in storehouse, i.e. pours position in storehouse layout work;
Put in storage machinery scheduler module: during the pouring construction of concrete dam, it usually needs multiple stage door machine, tower crane, cable
The concrete transporting equipments such as machine, tower belt crane, belt feeder, by concrete transportation to face, storehouse to complete follow-up work, the most just deposit
Multiple stage put in storage machinery scheduling problem;Corresponding with position in storehouse layout, in the scheduling process of concrete warehousing machinery, need to examine
Consider the parameters such as its conevying efficiency, transport capacity, locus, brachium, positive and negative lift, bulk, the domain of walk, and dam body
The phase relation column constraint of concreting;The concrete dam construction intelligent decision system of KBE can be according to working as
The integrated information of front job site each side, is meeting associated construction specification, dam concrete pouring technique requirement, progress and shape
On the premise of schedule requirement, inference system can arrange each concreting machinery that puts in storage to be responsible for the position in storehouse that pours successively, and
The method of operation of every set concrete warehousing equipment;
Pouring plan authentication module: the concrete dam construction scheme that system outer (and system user) is inputted, bag
Include and pour schedule, position in storehouse layout scheme, equipment scheduling scheme etc., the concrete dam construction intelligence of KBE
Decision system can be according to the integrated information of current job site each side, and meeting associated construction specification, dam concrete waters
On the premise of building technology requirement, progress and image schedule requirement, verify the feasibility of this pouring plan, as infeasible, be given former
Cause and explanation;
Construction speed (process) analog module: relevant information based on current job site comprehensive condition, knowledge based work
The concrete dam construction intelligent decision system of journey can meet associated construction specification, dam concrete pouring technique want
Ask, on the premise of progress and image schedule requirement, formulate the dam concrete in following certain period of time pour schedule or
Person remains the schedule that dam concrete pours, and the corresponding plan of getting the raw materials ready, grouting plan and water flowing cooling scheme,
Realize the artificial intelligence to concrete pouring construction progress;
Get the raw materials ready schedule module: the concrete dam construction intelligent decision system of KBE can be intended at native system
On the basis of fixed construction schedule, in conjunction with the concrete mix parameter at each position of dam body, reasoning and calculation obtains not
Carry out the demand of all types of cement (according to parameter classified types) in the construction time, aggregate, additive etc., for construction management with certainly
Plan uses;
Dam grouting schedule module: according to the practice of construction progress of concrete dam, the most each position in storehouse pour deadline, dam
Body pours the associated construction parameter such as cope level, upper pond level, according to associated specifications and requirement, mixing of KBE
Solidifying earth dam pouring construction intelligent decision system can obtain the position of dam body joint grouting with reasoning, recommend the construction ginsengs such as time started
Number, for construction management and decision references and;
Cooling water flowing scheme module: according to the practice of construction progress of concrete dam, the most each position in storehouse pour the deadline and
The hydrometeorological data of dam site location, according to associated specifications and requirement, the concrete dam pouring of KBE
When construction intelligent decision system can obtain the start-stop of the initial stage water flowing in each district of dam body, water flowing in mid-term and later stage water flowing with reasoning
Between, the parameter such as water temperature, flow, for construction management and decision-making.
See Fig. 1-6, the decision method of the concrete dam construction intelligent decision system of a kind of KBE, it
Comprise the steps:
(1) conventionally form represents the premise of knowledge and the extraction of conclusion,
In current concrete dam construction management with decision making process, whether a certain position in storehouse concrete can pour needs
Meet the restriction of before and after's discrepancy in elevation, the left and right discrepancy in elevation limits, the interlayer intermittent time limits, meteorological condition limits, adjacent position in storehouse is constructed about
Restraint these to limit and constraints;When all of precondition meets, this position in storehouse can carry out pouring construction;From knowledge work
The production rule angle of journey understands, with constraints, these limit can regard whether a certain position in storehouse can pour this rule as
Precondition, " can pour " is then corresponding to the conclusion of this rule;
In the step for knowledge is extracted, it is whether the concrete position in storehouse represented with conventionally form can pour this
One judges that work is converted into the premise part of rule, conclusion part, reliability form;Before whether concrete position in storehouse can pour
Before and after's discrepancy in elevation, the left and right discrepancy in elevation, intermittent time, meteorological condition, 5 aspects of adjacent construction are put forward;Conclusion is for pouring or can not
To pour, for 0-1 variable;According to concrete dam construction specification, this rule is completely reliable;
(2) rule premise, conclusion, the quantization of reliability, classification,
On the basis of (one) step work, carry out quantifying, classifying or classification by the premise of rule, conclusion, reliability;
The discrepancy in elevation front and back: represent with absolute physical value, i.e. represents with the actual discrepancy in elevation of front and back's position in storehouse, passes through Formal Languages
Be expressed as Height_Diff-FB_Bin (x, y);The discrepancy in elevation before and after i.e. position in storehouse x and position in storehouse y, position in storehouse x higher than position in storehouse y be on the occasion of,
Otherwise it is then negative value;
The left and right discrepancy in elevation: represent with absolute physical value, i.e. represents with the actual discrepancy in elevation of left and right position in storehouse, passes through Formal Languages
Be expressed as Height_Diff-LR_Bin (x, y);The i.e. left and right discrepancy in elevation of position in storehouse x and position in storehouse y, position in storehouse x higher than position in storehouse y be on the occasion of,
Otherwise it is then negative value;
Intermittent time: represent with absolute physical value, both poured the deadline plus rule with the bottom position in storehouse of current position in storehouse
On the basis of the intermittent time that model requires or current engineering sets, it is judged that whether current time had arrived in this time period;Pass through
Formal Languages is represented by Pause_time (x), and its value is absolute calendar time, its form be " MM/dd/yyyy hh:
mm;ss”;
Meteorological condition: whether current position in storehouse can be poured by meteorological condition mainly considers from rainfall angle, can be right
Rainfall carries out classification, as heavy rain, heavy rain, light rain, without rain;Being represented by Rain (t) by Formal Languages, " t " i.e. herein
For drafting the time of moduling plan, the value of Rain (t) is according to the weather forecast parameter determination in data base;
Adjacent position in storehouse construction constraint: the construction constraint of adjacent position in storehouse refers to when certain position in storehouse casting concrete, adjacent thereto
Other positions in storehouse because draining, template erection and the cause of failures that puts in storage cannot carry out concrete pouring construction simultaneously;This is constrained to
0-1 variable bound, when there are adjacent position in storehouse just at casting concrete, this position in storehouse can not casting concrete, this binding occurrence is
0;Otherwise, all positions in storehouse adjacent thereto are not the most at casting concrete, then current position in storehouse can carry out concrete pouring construction, should
Binding occurrence is 1;It is represented by Adjoin_Bin (x) by Formal Languages;Which position in storehouse to be the adjacent storehouse of current position in storehouse as
Position, determines according to the three-dimensional relationship of dam body model parameter;
Conclusion: whether can pour the judgement of sets of conditions according to above current position in storehouse, conclusion is for pouring or can not
To pour, for 0-1 variable;By Formal Languages, FitPour-Bin (x) can be expressed as;
Reliability: according to concrete dam construction specification, this rule is completely reliable, i.e. reliability is 1:CF=1;
(3) formalization of production rule,
In the concrete dam construction intelligent decision system of KBE, before rule in (two) step
Carry, the quantization of conclusion, reliability, classification, and the statement of Formal Languages, " during concrete dam construction certain
Whether one position in storehouse concrete can pour under current state " judgement of this event can be expressed as following production rule
Then:
The precondition of this rule be simultaneously meet before and after the discrepancy in elevation, the left and right discrepancy in elevation, intermittent time, meteorological condition, adjacent storehouse
Position construction constraint, it may be assumed that P=Height_Diff-FB_Bin (x, y) ∧ Height_Diff-LR_Bin (x, y) ∧ Pause_time
(x)∧Rain(t)∧Adjoin_Bin(x);The conclusion of this rule is that current position in storehouse can be with casting concrete, it may be assumed that C=
FitPour-Bin(x);The credibility of rule is 1;So, current rule can be expressed as:
Height_Diff-FB_Bin(x,y)∧Height_Diff-LR_Bin(x,y)∧Pause_time(x)∧Rain
(t)∧Adjoin_Bin(x)→FitPour-Bin(x)(1)
(4) rule inspection,
The inspection of rule is the important step of production rule acquisition work, be the examination to abovementioned steps gained knowledge and
Check on;In this step, major tasks are that by reflecting that this rule is verified by the example of various condition of construction, look into
See that its conclusion exported is the most correct;For the output incorrect situation of conclusion, need check its reasoning process in detail and find
Reason;In rule inspection, it is accordingly required in particular to verify that the adaptability of this rule is with correct by some construction parameters in particular cases
Property;If after system test, do not find current rule existing problems, then this rule warehouse-in is preserved;If it find that exist
Problem, then need again to extract current rule and formalization, i.e. repeats step ()-step (three);
(5) again extract or formalization, for being found to have the rule of reasoning impropriety in checkout procedure, repeat step
Suddenly (one)-step (three), extract or formalization again to this rule;
(6) warehouse-in, for rule qualified after inspection, is deposited in rule base, with the form of relational database
Preserve.
Illustrate the detailed description of the invention of the present invention in conjunction with Figure of description, mainly include the coagulation of a kind of KBE
Earth dam pouring construction intelligent decision system and the system constructing flow process of method, knowledge acquisition method, knowledge representation method, knowledge push away
Reason method.It should be noted that the process described by this place and embodiment not limitation of the present invention, but in order to more preferably
Ground describes present invention, is easy to the awareness and understanding present invention.
One, system constructing flow process
Fig. 2 is the building process schematic diagram of the concrete dam construction intelligent decision system of a kind of KBE.
Its building process is divided into four steps, respectively builds data base, builds knowledge base, inference system design, the system integration and tune
Examination.The main contents of each step are as follows:
(1) data base is built
Data base is the basis of the concrete dam construction intelligent decision system operation of a kind of KBE.Summarize
For, the process building data base is i.e. to affect the construction machinery of concrete dam construction, dam body feature, construction ring
The related datas such as border, technology requirement are sorted out, arrange, are quantified, and store through database table, through RDBM Relational Data Base Management system
System realizes the management to data base and maintenance.The concrete dam of a kind of KBE of the particular content of data base such as Fig. 3 waters
Build shown in the engineering information library structure of construction intelligent decision system.
(2) knowledge base is built
Knowledge base is the foundation that the concrete dam construction intelligent decision system of a kind of KBE makes inferences
Place.The building process of knowledge base is divided into the work of two aspects, an aspect to be relevant during concrete dam construction
The foundation of knowledge (i.e. rule) method for expressing, i.e. sets up concrete dam construction expressing for knowledge method;Another aspect is then
It is the knowledge that will preserve with traditional media or represent, including engineering discipline, technology requirement, expertise etc., is built with the present invention
Vertical mode shows, and i.e. sets up the acquisition methods of concrete dam construction knowledge.The particular content of knowledge base such as Fig. 5 mono-
Shown in the construction rule base structure of the concrete dam construction intelligent decision system planting KBE.
The content of these two aspects will elaborate in " knowledge representation method " and " knowledge acquisition method " hereinafter.
(3) inference system design
Inference system design is on the basis of constructed data base and knowledge base, it is achieved from basic data, lead to
Crossing the application of knowledge in knowledge base, reasoning obtains following construction schedule or the reasoning process of arrangement and method for construction.Inference system sets
Meter includes RBR uncertain inference machine-processed, rule-based mechanism, RBR conflict resolution plan
Slightly three aspect contents, its particular content will elaborate in " Method of Knowledge Reasoning " in " detailed description of the invention ".
(4) system integration and debugging
After the design of the structure and inference system that complete data base and knowledge base, carry out the integrated of system and adjust
Examination.The step for purpose be the linking between real data storehouse, knowledge base, inference system three, and make this three work in coordination with
Work, it is achieved the reasoning from concrete dam construction basic data to arrangement and method for construction.A kind of concrete of KBE
The system integration of dam pouring construction intelligent decision system and the major tasks of debugging include the content of three aspects: one is data base
Can the data inside comprised provide precondition, the i.e. completeness of production rule precondition for knowledge base;Two is rule
Between conflicting, i.e. rule concordance, this partial content will be explained in the Strategy of Conflict Resolution of Process Based in detail
State;Three is the explanation of reasoning process, and this partial content will elaborate in RBR designs.
Complete above quadrinomial job, i.e. achieve the concrete dam construction intelligent decision of a kind of KBE
The building process of system, after the real data in input concrete dam engineering foundation data and work progress, can be carried out
The management of concrete dam construction and decision-making work.
Two, knowledge acquisition method
The product of the concrete dam construction intelligent decision system of KBE as a kind of in the Fig. 4 in Figure of description
Shown in raw formula Rule schematic flow sheet, obtain concrete dam construction dependency rule and be broadly divided into following six aspect work
Make.
Below with a certain position in storehouse concrete during concrete dam construction whether can pour under current state into
Example, illustrates that the knowledge of concrete dam construction extracts process.
(1) conventionally form represents the premise of knowledge and the extraction of conclusion
In current concrete dam construction management with decision making process, whether a certain position in storehouse concrete can pour needs
Meet the restriction of before and after's discrepancy in elevation, the left and right discrepancy in elevation limits, the interlayer intermittent time limits, meteorological condition limits, adjacent position in storehouse is constructed about
Restraint these to limit and constraints.I.e. when all of precondition meets, this position in storehouse can carry out pouring construction.From knowledge
The production rule angle of engineering understands, with constraints, these limit can regard whether a certain position in storehouse can pour these rule as
Precondition then, " can pour " is then corresponding to the conclusion of this rule.
In the step for knowledge is extracted, focus is whether may be used the concrete position in storehouse that represent with conventionally form
With pour this judgement work be converted into rule premise part, conclusion part, reliability form.Concrete position in storehouse is the most permissible
The premise poured has before and after's discrepancy in elevation, the left and right discrepancy in elevation, intermittent time, meteorological condition, 5 aspects of adjacent construction;Conclusion is for watering
Build and maybe cannot pour, for 0-1 variable;According to concrete dam construction specification, this rule is completely reliable.
(2) rule premise, conclusion, the quantization of reliability, classification
On the basis of the first step works, carry out quantifying, classifying or classification by the premise of rule, conclusion, reliability.Above
State as a example by the most whether a certain position in storehouse in concrete dam construction can pour this problem, elaborate this work
Implementation.
The discrepancy in elevation front and back: represent with absolute physical value, i.e. represents with the actual discrepancy in elevation of front and back's position in storehouse, passes through Formal Languages
Be expressed as Height_Diff-FB_Bin (x, y).The discrepancy in elevation before and after i.e. position in storehouse x and position in storehouse y, position in storehouse x higher than position in storehouse y be on the occasion of,
Otherwise it is then negative value.
The left and right discrepancy in elevation: represent with absolute physical value, i.e. represents with the actual discrepancy in elevation of left and right position in storehouse, passes through Formal Languages
Be expressed as Height_Diff-LR_Bin (x, y).The i.e. left and right discrepancy in elevation of position in storehouse x and position in storehouse y, position in storehouse x higher than position in storehouse y be on the occasion of,
Otherwise it is then negative value.
Intermittent time: represent with absolute physical value, both poured the deadline plus rule with the bottom position in storehouse of current position in storehouse
On the basis of the intermittent time that model requires or current engineering sets, it is judged that whether current time had arrived in this time period.Pass through
Formal Languages is represented by Pause_time (x), and its value is absolute calendar time, its form be " MM/dd/yyyy hh:
mm;ss”。
Meteorological condition: whether current position in storehouse can be poured by meteorological condition mainly considers from rainfall angle, can be right
Rainfall carries out classification, as heavy rain, heavy rain, light rain, without rain.Being represented by Rain (t) by Formal Languages, " t " i.e. herein
For drafting the time of moduling plan, the value of Rain (t) is according to the weather forecast parameter determination in data base.In concrete dam work
In journey Construction practices, the concrete of different performance is different to the requirement of rainfall;Additionally, with or without rainproof measure or rainproof measure
Degree of perfection also has different impacts to this judgement, sweetly disposition in concrete engineering, and arranges corresponding Formal Languages
As basis for estimation.
Adjacent position in storehouse construction constraint: the construction constraint of adjacent position in storehouse refers to when certain position in storehouse casting concrete, adjacent thereto
Other positions in storehouse (elevation is the most identical) because draining, template erection, the reasons such as interference of putting in storage cannot carry out concrete simultaneously
Pouring construction.This is constrained to 0-1 variable bound, and when there are adjacent position in storehouse just at casting concrete, this position in storehouse can not pour
Concrete, this binding occurrence is 0;Otherwise, all positions in storehouse adjacent thereto are not the most at casting concrete, then current position in storehouse can be carried out
Concrete pouring construction, this binding occurrence is 1.It is represented by Adjoin_Bin (x) by Formal Languages.As for which position in storehouse it is
The adjacent position in storehouse of current position in storehouse, determines according to the three-dimensional relationship of dam body model parameter.
Conclusion: whether can pour the judgement of sets of conditions according to above current position in storehouse, conclusion is for pouring or can not
To pour, for 0-1 variable.By Formal Languages, FitPour-Bin (x) can be expressed as.Require emphasis a bit herein, this
According to the integrated condition of position in storehouse itself, one rule simply judges whether current position in storehouse can pour, do not necessarily mean that and necessarily may be used
To pour or have to pour.
Reliability: according to concrete dam construction specification, this rule is completely reliable, i.e. reliability is 1:CF=1.
(3) formalization of production rule
In the concrete dam construction intelligent decision system of KBE, before rule in second step
Carry, the quantization of conclusion, reliability, classification, and the statement of Formal Languages, " during concrete dam construction certain
Whether one position in storehouse concrete can pour under current state " judgement of this event can be expressed as following production rule
Then:
The precondition of this rule be simultaneously meet before and after the discrepancy in elevation, the left and right discrepancy in elevation, intermittent time, meteorological condition, adjacent storehouse
Position construction constraint, it may be assumed that P=Height_Diff-FB_Bin (x, y) ∧ Height_Diff-LR_Bin (x, y) ∧ Pause_time
(x)∧Rain(t)∧Adjoin_Bi n(x).The conclusion of this rule is that current position in storehouse can be with casting concrete, it may be assumed that C=
FitPour-Bin(x).The credibility of rule is 1.So, current rule can be expressed as:
Height_Diff-FB_Bin(x,y)∧Height_Diff-LR_Bin(x,y)∧Pause_time(x)∧Rain
(t)∧Adjoin_Bin(x)→FitPour-Bin(x)(1)
(4) rule inspection
The inspection of rule is the important step of production rule acquisition work, be the examination to abovementioned steps gained knowledge and
Check on.In this step, major tasks are that by reflecting that this rule is verified by the example of various condition of construction, look into
See that its conclusion exported is the most correct.For the output incorrect situation of conclusion, need check its reasoning process in detail and find
Reason.In rule inspection, it is accordingly required in particular to verify that the adaptability of this rule is with correct by some construction parameters in particular cases
Property.If after system test, do not find current rule existing problems, then this rule warehouse-in is preserved;If it find that exist
Problem, then need again to extract current rule and formalization, i.e. repeats step 1-step 3.
(5) again extract or formalization
For being found to have the rule of reasoning impropriety in checkout procedure, repeat step 1-step 3, this rule is carried out
Again extract or formalization.
(6) warehouse-in
For rule qualified after inspection, being deposited in rule base, concrete production rule is in rule base
Storage mode i.e. knowledge representation method will describe in detail in the Part III " knowledge representation method " of " detailed description of the invention ".
Three, knowledge representation method
As a example by whether certain concrete position in storehouse can pour, elaborate the concrete dam pouring of a kind of KBE
The knowledge representation method of construction intelligent decision system, i.e. the method for expressing of each rule-like of concrete dam construction.
As it has been described above, concrete dam construction process control rule (i.e. knowledge) is Activities during construction management
Between causal conversion foundation, applicable production rule represents.Production rule is to represent with " P → C (CF) " form
Statement.Data for utilizing ripe relational database technology powerful stores, retrieve, organize and operating capability realization is to rule
The organization and management of knowledge, utilizes in the concrete dam construction intelligent decision system of a kind of KBE and closes coefficient
According to library storage, management rule.All of rule in the concrete dam construction intelligent decision system of a kind of KBE
It is saved in rule base (Rule Base).According to the feature of concrete dam construction dependency rule, concrete dam pouring is advised
Relational database representation method then has carried out detailed design.
Concrete dam pouring rule base is divided into dictionary word bank (Dictionary sub-base) and rule word bank (Rule
Sub-base) two parts.Dictionary word bank for deposit rule precondition and conclusion (being referred to as the fact), and the fact from
The contents such as right language description.The structure of dictionary word bank is DICT_SB (Fact_ID, Fact, Nat_lang, CF).Wherein Fact_
ID is true coding, and the fact that Fact is to describe with formal language content, the fact that Nat_lang is to use natural language description
Content (for explaining to user), CF refers to the credibility of the fact, is all set at the CF of all facts of reasoning original state
Empty.Wherein, Fact_ID is major key.Dictionary in the concrete dam construction intelligent decision system of a kind of KBE
The fragment in storehouse is as shown in table 1.
Table 1 dictionary word bank fragment
Illustrate:
(1) table 1 data are only the fragment of dictionary character library of rule base, and the regular quantity required when running of real system is relatively
For huge and complicated.
(2) the part fact in table 1 is reflected as functional relationship, and the realization of its function need to rely on the set up system of the present invention
Other functional modules in system.Such as: the neighbouring judgement of position in storehouse, the judgement of the judgement of left and right position in storehouse, front and back position in storehouse, it is required to borrow
Help three-dimensional dam body model and spatial parameter determines.
(3) fact of the part in table 1 needs to make inferences with the data in field data storehouse for relying in reasoning process.
Rule word bank is used for depositing rule, including regular texture table and rule condition table.The structure of regular texture table is
RULE_STRU(Rule_ID,Num,Fact_ID,CF,Active).Wherein Rule_ID is rule numbers;Num is this rule institute
The condition item number needed;Fact_ID is the numbering of conclusion;CF is the credibility of rule;Active shows that this rule is the most
Activate, it is to avoid reusing of rule, improve Reasoning Efficiency.The structure of rule condition table be RULE_Cond (Rule_ID,
Fact_ID,Logic_relation,Active).Wherein Rule_ID represents rule numbers;Fact_ID represents condition number, should
Rule has several condition just to have several records;Logic_relation is this condition logical relation in reasoning, including patrolling
Volume with " ∧ " (" AND ") computing and logic or " ∨ " (" OR ") two kinds of operational forms of computing, represent that this condition is essential condition
Or optional condition;Active is the mark that Fact_ID activates, it is to avoid rule terms and conditions repeated matching.
The CF(credibility of rule) refer to the degree of reliability that foundation rule condition is reached a conclusion.CF (P → C) ∈ [-1,1],
Value and the implication of CF are as follows: CF=+1 represents that C mono-is set to very if P is true;CF=-1 represents that C mono-is set to vacation if P is true;
CF=0 represents that C is not affected by the true and false of P;CF ∈ (-1,0) represents if P is true, and C is false to a certain extent;CF∈(0,
1) represent that, if P is true, C is true to a certain extent.
Regular texture table in rule word bank and the fragment of rule condition table respectively as shown in table 2, table 3, the fragment in table
Reflect the most whether certain position in storehouse can carry out this construction rule of concreting.
Table 2 regular texture table fragment
Rule_ID |
Num |
Fact_ID |
CF |
Active |
R0001 |
1 |
F0025 |
1 |
F |
R0002 |
2 |
F0032 |
0.9 |
F |
R0008 |
5 |
F0010 |
1 |
F |
… |
… |
… |
… |
|
Table 3 rule condition table fragment
Rule_ID |
Fact_ID |
Logic_relation |
Active |
R0008 |
F0002 |
AND |
F |
R0008 |
F0003 |
AND |
F |
R0008 |
F0006 |
AND |
F |
R0008 |
F0007 |
AND |
F |
R0008 |
F0009 |
AND |
F |
… |
… |
|
|
There is above agreement, it is possible to the rule in the rule represented at natural language and relational database mutually turns
Change.According to table 2, table 3, the form of rule R0008 is as follows:
Height_Diff-FB_Bin(x,y)∧Height_Diff-LR_Bin(x,y)∧Pause_time(x)∧Rain
(t)∧Adjoin_Bin(x)→FitPour-Bin(x)(1)
Again according to table 1, rule R0008 can be explained further into: if a certain position in storehouse " x " meet simultaneously before and after the discrepancy in elevation, a left side
The right discrepancy in elevation, intermittent time, meteorological condition, adjacent position in storehouse are constructed 5 item constraints, then this position in storehouse can carry out concreting, rule
Reliability be " 1 ".Thus establish concrete dam construction specification and expertise knowledge and computer can be grasped
One-to-one relationship between the rule knowledge made, it is achieved that the intelligence in the computer information system of rule knowledge represents and deposits
Storage.
Four, Method of Knowledge Reasoning
Knowledge reasoning refers to from existing basic data, by answering various production rules and the comprehensive of knowledge
With, obtain the current decision-making of system or action scheme through reasoning.Specific to concrete dam construction problem, refer to according to mixed
Coagulate the current execution conditions of earth dam and construction environment, before the related specifications and technology following concrete dam construction requires
Putting, reasoning obtains the process of current concrete dam construction scheme.
The concrete dam construction intelligent decision system of a kind of KBE and the Method of Knowledge Reasoning master of method
Disappear including rule-based uncertain inference mechanism, RBR Model Design, the conflict of Process Based
Solve three aspect contents of strategy.
The fact that during concrete dam construction and rule all exists uncertainty, with concrete dam construction
Based on the uncertain true expression regular with uncertainty and memory mechanism, set up concrete dam construction rule-based
Uncertain inference mechanism, it is achieved the uncertain inference of concrete dam pouring system, i.e. intelligent inference.
The uncertain inference mechanism of the concrete dam construction intelligent decision system of a kind of KBE uses
C-F uncertain inference model, C-F model be uncertain inference uses the earliest, simple and effective a kind of inference method.
A kind of uncertain inference algorithm bag in the concrete dam construction intelligent decision system of KBE
Include four parts, ask the credibility of conclusion, the same knot of two independent premise derivation including according to the credibility of premise and rule
The merging of the credibility of opinion, the conjunction of premise, the extracting of premise, the most just can carry out various reasoning computing.
Wherein CF (P) represents the credibility of premise;The credibility of CF (R) delegate rules;CF (C) represents the credibility of conclusion.
(1) credibility of conclusion is sought according to the credibility of premise and rule
CF(C)=CF(R)×max{0,CF(P)}
If CF (P) > 0, i.e. rule premise to be true to a certain degree, then calculates rule according to CF (P) and Rules control CF (R)
Then credibility CF (C) of conclusion=CF (R) × CF (P);
If CF (P)=0, i.e. condition are true or false and know nothing, then CF (C)=0, say, that conclusion true and false also one without institute
Know.
If CF (P) < 0, i.e. rule premise are false, illustrate that this rule can not be applied, then CF (C)=0.
The merging of the credibility of the same conclusion of (2) two independent premise derivation, also referred to as fresh evidence (premise) rule
When combining the credibility of the same conclusion that plural independent premise derives, first merge two of which, its knot
Fruit merges with the 3rd again, and the rest may be inferred.
(3) conjunction of premise
For P=P1∧P2∧P3∧…∧Pn, have
CF(P)=CF(P1∧P2∧P3∧…∧Pn)=min{CF(P1),CF(P2),…,CF(Pn)}
From above formula, the CF of the conjunction of multiple premises is the CF of that minimum for CF premise.
(4) the extracting of premise
For P=P1∨P2∨P3∨…∨Pn, have
CF(P)=CF(P1∨P2∨P3∨…∨Pn)=max{CF(P1),CF(P2),…,CF(Pn)}
From above formula, the CF extracted of multiple premises is the CF of that maximum for CF premise.
For uncertain inference, when the credibility of certain conclusion is not 1, for same conclusions other rules still
Make inferences, obtain the credibility of conclusion, and merge with the credibility of this conclusion obtained.
A kind of concrete dam construction intelligent decision system of KBE using uncertain inference rule as
Meta-knoeledge, is individually placed in reasoning module, is allowed to have automatic focusing properties, i.e. priority level the highest.This ensures that there two
The rule (having different credibility) with same conclusions had been combined with before being triggered by other rules.
If Fig. 6 is reasoning pattern and the work of the concrete dam construction intelligent decision system of a kind of KBE
Shown in flow process, the inference machine of the concrete dam construction intelligent decision system of a kind of KBE is according in data base
Data, utilize the rule in knowledge base, by certain inference strategy, go Solve problems, the data explaining outside input and the fact,
Derive conclusion and explain to user, the reason reached a conclusion is described as requested.
The concrete dam construction intelligent decision system of a kind of KBE belongs to design, the knowledge system of planning class
System, solution space is very big, so selecting forward reasoning mode.
Forward reasoning is also referred to as data drive control, is from known data message, and forward uses rule, allows rule
Former piece match with the record in data base and the fact, then enable rule conclusion part.
First forward reasoning is provided master data by user, leaves in data base, and then inference machine is operated, and it pushes away
The key step of reason process is as follows:
(1) inference machine mates with the former piece of rule in rule base by this batch of fact.If the rule of only one coupling
Then, then this rule is selected;If there being many rules to use, then select a rule therein according to Strategy of Conflict Resolution.
(2) the conclusion part fact of this rule is added in dynamic data base as new fact.
(3) with the fact that in the data base after renewal, (1), (2) step are repeated.So the most repeatedly until Resolving probiems or be given
Till indeterminable reason.
So-called conflict refers to that a certain step in concrete dam construction solution formulation reasoning process has many rules can
To use.Conflict resolution refers to select one from many rules that these are available for current reasoning;Strategy of Conflict Resolution
It is how to select strategy or the method for best rule from many rules.
Feature according to concrete dam construction process: along with the increase of rule pre-condition, conclusion is currently
Under the conditions of specific aim strengthen;Rules control is the highest, and conclusion is the most reliable.So at the concrete dam of KBE
During pouring construction solution formulation, during rule-based uncertain inference, use following Strategy of Conflict Resolution.
Assume R1、R2For current spendable rule, then:
①GetNum(R1)、Num(R2)、CF(R1)、CF(R2)
②if Num(R1)Num(R2)then first_use_rule=R2goto⑤
③if Num(R2)Num(R1)then first_use_rule=R1goto⑤
④if CF(R1)≤CF(R2)then first_use_rule=R2else first_use_rule=R1
⑤end
Wherein Num (R1)、Num(R2) it is R1、R2Condition set, CF (R1)、CF(R2) it is R1、R2Credibility.When there being more than two
Rule when can use, try to achieve first with this strategy and front two rules preferentially use rule, then this rule is advised with Article 3
Then compare, carry out successively, until finding best rule.