CN108388666A - A kind of database multi-list Connection inquiring optimization method based on glowworm swarm algorithm - Google Patents
A kind of database multi-list Connection inquiring optimization method based on glowworm swarm algorithm Download PDFInfo
- Publication number
- CN108388666A CN108388666A CN201810216842.9A CN201810216842A CN108388666A CN 108388666 A CN108388666 A CN 108388666A CN 201810216842 A CN201810216842 A CN 201810216842A CN 108388666 A CN108388666 A CN 108388666A
- Authority
- CN
- China
- Prior art keywords
- firefly
- algorithm
- database
- swarm algorithm
- glowworm swarm
- 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
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/24—Querying
- G06F16/245—Query processing
- G06F16/2453—Query optimisation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
Abstract
The present invention is suitable for database field more particularly to a kind of database multi-list Connection inquiring optimization method based on glowworm swarm algorithm.The method includes:Database multi-list Connection inquiring optimization Cost Model is established in the search space for defining database;Attended operation in the solution space of left deep tree composition is encoded;Define fitness function;Glowworm swarm algorithm is introduced, each parameter is initialized, initializes firefly position and brightness;According to the attraction rule between firefly brightness size and firefly, weighting function and adaptive step mechanism are introduced, the update of all firefly positions and brightness is completed, best queries executive plan when database multi-list connection is searched out by glowworm swarm algorithm.The present invention provides a kind of database multi-list Connection inquiring optimization method based on glowworm swarm algorithm, and best queries executive plan when searching out database multi-list connection by glowworm swarm algorithm improves database multi-list search efficiency by execution best queries executive plan.
Description
Technical field
The present invention relates to database field more particularly to a kind of database multi-list Connection inquiring based on glowworm swarm algorithm are excellent
Change method.
Background technology
With the appearance of large scale database and data warehouse, database size increasingly increases, and data query frequency increases,
How to find and meet user's search request scheme, improves Query Efficiency, become the hot issue of current research.
Query optimization is the basis of database application, and multi-join query optimization problem is a np problem, but cannot all be protected
Card can provide an optimal executive plan within the limited time.It is low for Query Efficiency, it is difficult to which that acquisition is most preferably looked into
The limitation of prioritization scheme is ask, heuritic approach is introduced into data base querying by some scholars.
Currently, in database inquiry optimization method, usually used method has:1. traditional exhaustive search algorithm 2. is true
Algorithm is determined, such as dynamic programming algorithm, greedy algorithm;3. Stochastic Optimization Algorithms, such as genetic algorithm, particle cluster algorithm, ant group algorithm
Deng.
Existing method for transformation is generally designated as there are certain defect:
1. traditional exhaustive search algorithm can not to comprising the too big inquiry of relationship number optimize.
2. determining in algorithm, dynamic programming algorithm is not suitable for the optimization of complicated multi-join query, and greedy algorithm can only
Acquire the locally optimal solution in current meaning.
3. in Stochastic Optimization Algorithms, it cannot be guaranteed that one surely obtains the optimal solution of problem, but approximate optimal solution can be obtained, it can
To improve query optimization efficiency, but it is slow in the algorithm later stage to be likely to occur convergence rate, is easily absorbed in locally optimal solution etc..
The optimization of database multi-list Connection inquiring is to improve the key technology of database performance, is connected to improve database multi-list
Connect search efficiency, it is proposed that a kind of database multi-list Connection inquiring optimization method based on glowworm swarm algorithm is calculated by firefly
Method finds best queries executive plan, improves database multi-list Connection inquiring efficiency.
Invention content
The major function of database is storage and management data, provides a user query function.Connect for database multi-list
It is low to connect search efficiency, the problem of algorithm later stage is easily absorbed in locally optimal solution, the present invention provides a kind of number based on glowworm swarm algorithm
According to library multi-table join enquiring and optimizing method, best queries executive plan is found by glowworm swarm algorithm, data base querying is executed
Plan optimizes, and improves database multi-list Connection inquiring efficiency.
To achieve the goals above, present invention employs following technical solutions:
1. defining the search space of database, database multi-list Connection inquiring optimization Cost Model is established;
2. in the solution space of left deep tree composition, coded sequence is obtained using follow-up traversal threaded tree;
3. defining fitness function;
4. application glowworm swarm algorithm finds best queries executive plan;
5. meeting termination condition, best queries executive plan is exported.
Wherein, best queries executive plan is found using glowworm swarm algorithm, algorithm steps include in the present invention:
(1) initialization algorithm basic parameter, setting firefly number n, maximum Attraction Degree β0, light intensity absorption coefficient gamma, step
Long factor-alpha, maximum iteration MaxGeneration or search precision ε;
(2) position of random initializtion firefly, according to fitness function calculate firefly fitness value as it most
Big fluorescence intensity I0;
(3) the relative luminance I and Attraction Degree β for calculating firefly compare the fluorescent brightness size of firefly in affiliated neighborhood,
The moving direction of firefly is determined according to relative luminance;
(4) according to the linear decrease weighting function of introducing and introducing adaptive equalization mechanism, the position of firefly is updated;
(5) part may be absorbed in most if optimal value does not update continuously three times to the firefly for being in optimum position
It is excellent, then random perturbation is carried out to the position of optimal firefly, the random perturbation of a Gaussian distributed is added, calculation can be made
Method jumps out local optimum;
(6) according to the position of firefly after update, the brightness of firefly is recalculated;
(7) global extremum point and optimum individual value are exported if meeting end condition, the corresponding number in optimal firefly position
According to library best queries executive plan, otherwise, searching times increase by 1, turn (3) step, are searched for next time.
Description of the drawings
Fig. 1 is a kind of database multi-list Connection inquiring optimization method based on glowworm swarm algorithm provided in an embodiment of the present invention
Flow chart;
Fig. 2 is the algorithm flow that a kind of glowworm swarm algorithm provided in an embodiment of the present invention finds best queries executive plan
Figure.
Specific implementation mode
In order to make the purpose , technical scheme and advantage of the present invention be clearer, hereinafter reference will be made to the drawings and it is real to combine
The invention will be further described for example, it should be understood that and the specific embodiments described herein are merely illustrative of the present invention, and
It is not used in the restriction present invention.
In the embodiment of the present invention, the search space of database is defined, establishes database inquiry optimization Cost Model;In left depth
In the solution space for setting composition, coded sequence is obtained using follow-up traversal threaded tree;Define fitness function;It is calculated using firefly
Method finds best queries executive plan;Meet termination condition, exports best queries executive plan.Wherein, using glowworm swarm algorithm
Finding best queries executive plan algorithm steps includes:(1) firefly number n is arranged in initialization algorithm basic parameter, maximum
Attraction Degree β0, light intensity absorption coefficient gamma, step factor α, maximum iteration MaxGeneration or search precision ε;(2) with
Machine initializes the position of firefly, and the fitness value of firefly is calculated as its maximum fluorescence intensity I according to fitness function0;
(3) the relative luminance I and Attraction Degree β for calculating firefly compare the fluorescent brightness size of firefly in affiliated neighborhood, according to phase
To the moving direction of brightness decision firefly;(4) according to the linear decrease weighting function of introducing and introducing adaptive equalization machine
System, updates the position of firefly;It (5) can if optimal value does not update continuously three times to being in the firefly of optimum position
It can be absorbed in local optimum, then random perturbation is carried out to the position of optimal firefly, the random of Gaussian distributed is added and disturbs
It is dynamic, algorithm can be made to jump out local optimum;(6) according to the position of firefly after update, the brightness of firefly is recalculated;(7)
If meeting end condition, global extremum point and optimum individual value are exported, the corresponding database in optimal firefly position is most preferably looked into
Executive plan is ask, otherwise, searching times increase by 1, turn (3) step, are searched for next time.
In order to illustrate technical solutions according to the invention, illustrated below by specific embodiment.
Fig. 1 is a kind of database multi-list Connection inquiring optimization method based on glowworm swarm algorithm provided in an embodiment of the present invention
Flow chart.Referring to Fig.1, the embodiment of the present invention provides a kind of database multi-list Connection inquiring optimization based on glowworm swarm algorithm
Method, the method includes:
1. defining the search space of database, database inquiry optimization Cost Model is established;
2. in the solution space of left deep tree composition, coded sequence is obtained using follow-up traversal threaded tree;
3. defining fitness function;
4. application glowworm swarm algorithm finds best queries executive plan;
5. meeting termination condition, best queries executive plan is exported.
Wherein, it includes the following contents that step 4 application glowworm swarm algorithm, which finds best queries executive plan, with reference to Fig. 2.Fig. 2
For a kind of algorithm flow chart of glowworm swarm algorithm searching best queries executive plan provided in an embodiment of the present invention.Reference Fig. 2,
The embodiment of the present invention provides a kind of application glowworm swarm algorithm searching best queries executive plan method, and the method step includes:
(1) initialization algorithm basic parameter, setting firefly number n, maximum Attraction Degree β0, light intensity absorption coefficient gamma, step
Long factor-alpha, maximum iteration MaxGeneration or search precision ε;
(2) position of random initializtion firefly calculates the fitness value of firefly as it according to fitness function
Maximum fluorescence intensity I0;
(3) the relative luminance I and Attraction Degree β for calculating firefly compare the fluorescent brightness size of firefly in affiliated neighborhood,
The moving direction of firefly is determined according to relative luminance;
(4) according to the linear decrease weighting function of introducing and introducing adaptive equalization mechanism, the position of firefly is updated;
(5) part may be absorbed in most if optimal value does not update continuously three times to the firefly for being in optimum position
It is excellent, then random perturbation is carried out to the firefly of optimal location, the random perturbation of a Gaussian distributed is added, calculation can be made
Method jumps out local optimum;
(6) according to the position of firefly after update, the brightness of firefly is recalculated;
(7) global extremum point and optimum individual value are exported if meeting end condition, the corresponding number in optimal firefly position
According to library best queries executive plan, otherwise, searching times increase by 1, turn (3) step, are searched for next time.
In above-mentioned steps 1, the search space of database is defined, establishes database inquiry optimization Cost Model.One inquiry
Corresponding all query execution plans constitute the policy space of the inquiry, and the size of policy space is turned by the equivalence followed
The physical operations collection that rule set and query execution engine are supported is changed to determine.Since policy space is generally all very big, inquiry
Optimization algorithm will not usually scan in entire policy space, but in an Optimum Implementation Plan plan that may be present
It is slightly scanned in space subset, i.e., search space carries out.Because the left spaces Shen Shu are a fairly small sons of policy space
Collection, and the left spaces Shen Shu include usually best executive plan, at least also include preferable approximate optimal solution.And because
It is much smaller than the memory space needed for right deep tree for the memory space that the left spaces Shen Shu need, so frequently with preferably left depth
Space is set to simplify the solution of optimization problem.
Database multi-list can be related to multiple relation tables during inquiring, and the order of connection between different tables can cause to look into
Ask the diversity of plan.For n table, the left and right deep tree all factorial inquiry plan containing n, query optimizations are exactly to find one
The executive plan of item minimum Query Cost.
Assuming that one contains n relationship R={ r1,r2,…,rnThreaded tree, the internal node number of syntax tree is
ti, then the Executing Cost inquired is the summation of syntax tree internal node number.
For a multi-join query sentence, S is the inquiry plan that the query statement that all users provide has identical result
Set slightly, then the query statement of query execution Least-cost is exactly optimal inquiry plan in this set.
cos t(s0)=Min (cos t (s)), s ∈ S
Wherein,r1,r2,…,ri∈ R, i=1,2 ..., n-1.
In above-mentioned steps 2, in the solution space of left deep tree composition, coded sequence is obtained using follow-up traversal threaded tree.
For the present embodiment using left linear tree space as search space, database multi-join query optimization problem can be simple
Ground is described as:All vertex construct the binary tree of a minimum cost as leaf node using in vertex set V, and each vertex goes out
Existing and only appearance is primary.Based on left linear space multi-join query optimization problem, it is contemplated that query execution plan is in time and sky
Between on complexity, and its cost assessment depending on operating result collection record count, so by Connection inquiring optimization ask
Topic is reduced to the logic optimization problem of the join trees connection order of n table (relationship) in inquiry, and such multi-join query optimization is asked
Topic is converted to similar typical TSP problems.
In above-mentioned steps 3, fitness function is defined, it is in place with the absolute brightness characterization firefly institute of firefly in algorithm
Set the target function value at place.It is possible to reach or connect in optimization calculates to weigh each firefly in group using fitness
It is bordering on the excellent degree of optimal solution, to obtain the search information of next step, the function for measuring individual adaptation degree is exactly to adapt to letter
Number.
Fitness function can be divided into simple fitness function, linearly accelerate fitness function, non-linear acceleration fitness function and row
Sequence fitness function.It is both generally intended to use simple fitness function, also applies simple fitness function herein.
The convergent of the fitness function person's of directly affecting glowworm swarm algorithm.The target of query optimization is cost to be obtained
The executive plan of cost (t) minimums, therefore the fitness function of firefly individual must be related to cost cost (X), therefore it is suitable
Response value function is defined as follows:
The Executing Cost of firefly is calculated first, then inverted to obtained value is fitness function value.Work as firefly
The Executing Cost of fireworm is got over hour, and fitness is higher, then the effect inquired is better.
In above-mentioned steps 4, best queries executive plan is found using glowworm swarm algorithm, it can be with by introducing glowworm swarm algorithm
It finds database multi-list and connects best queries executive plan.
In above-mentioned steps (1), initialization algorithm basic parameter, setting firefly number n, maximum Attraction Degree β0, light intensity
Absorption coefficient γ, step factor α, maximum iteration MaxGeneration or search precision ε.
Wherein, β0For greatest attraction forces, i.e., at light source (at r=0) firefly attraction, take β0=1;
γ is the absorption coefficient of light, has indicated the variation of attraction, convergence rate and optimization of its value to glowworm swarm algorithm
Effect has a great impact, and for most problem, can take γ ∈ [0.01,100];
α is constant, can generally take α ∈ [0,1];
In the present embodiment, β is taken0=1.0, γ=1.0, α=1.0,Wherein rand is equal on [0,1]
The random number of even distribution.
In above-mentioned steps (2), the position of random initializtion firefly in solution room calculates firefly according to fitness function
The fitness value of fireworm is as its maximum fluorescence intensity Ii。
In the solution space, glowworm swarm algorithm randomly initializes a group firefly xi(i=1,2 ..., n), n is the light of firefly
The number of worm, xi=(xi1,xi2,…,xid) it is a D dimensional vector, indicate positions of the firefly i in solution space, Ke Yidai
One potential solution of the table problem.
Fitness value is calculated according to initial position, and as the maximum fluorescence intensity I of firefly0。
In above-mentioned steps (3), calculate the relative luminance I and Attraction Degree β of firefly, relatively belonging in neighborhood firefly it is glimmering
Brightness size determines the moving direction of firefly according to relative luminance.
Relative fluorescence brightness is:Wherein, I0Indicate i-th firefly maximum fluorescence brightness;λ is light
Intensity absorption coefficient;Degree of attracting each other β determines the distance of firefly movement, is defined as:
Wherein, β0Indicate maximum Attraction Degree, i.e. Attraction Degree at light source (at r=0).
rijFor the distance between firefly i and j, it is defined as:
In above-mentioned steps (4), according to the linear decrease weighting function of introducing and adaptive equalization mechanism is introduced, updates the light of firefly
The position of worm.
In glowworm swarm algorithm, due to being attracted by firefly i, firefly j is moved to it and is updated the position of oneself, j
It is as follows to set more new formula:
xj(t+1)=xj(t)+βij(rij)(xi(t)-xj(t))+αεj
In formula:T is the iterations of algorithm, xi, xjFor the spatial position residing for ith and jth firefly;α be step-length because
Son;Wherein rand is the equally distributed random number on [0,1].
The glowworm swarm algorithm of standard is in the algorithm iteration later stage, since firefly has been gradually moved into part or global pole
Near value point, the distance between firefly is gradually reduced at this time, and the Attraction Degree between firefly gradually increases, it will makes firefly
The displacement distance of fireworm individual is excessive and can not reach or miss optimal location, causes the problem of Near The Extreme Point shakes,
Therefore linear decrease weighting function is introduced in the present embodiment
Simultaneously because the firefly moving step length of algorithm design is fixed α, it is unfavorable for algorithm later stage solution local optimum
Value, and adaptive step adjustment is conducive to improve the precision and convergence rate solved.Therefore adaptive equalization mechanism, step are introduced
The constringency performance that long factor-alpha value influences glowworm swarm algorithm influences, and can further improve algorithm the convergence speed and global optimizing energy
Power uses Step-varied back propagation mechanism in searching process:Initial stage α value is larger, improves global optimizing ability;With iteration time
Number increases, and gradually reduces α values, accelerates late convergence, and specific adjustment mode is:α=α × Δ α, in formula, Δ α is step-length
Step-length attenuation coefficient, in (0.95,1) interior value.
Location formula is updated to:
xj(t+1)=w (t) xj(t)+βij(rij)(xi(t)-xj(t))+α×Δαεj
Wherein, x, I points are firefly position and brightness.
Influence of the firefly last time location information to current location can be controlled by inertia weight, weight size determines
Firefly movement apart from size, and strengthen the global optimizing and local search ability of glowworm swarm algorithm.
It,, may if optimal value does not update continuously three times to being in the firefly of optimum position in above-mentioned steps (5)
It is absorbed in local optimum, then random perturbation is carried out to the position of optimal firefly, the random of Gaussian distributed is added and disturbs
It is dynamic, algorithm can be made to jump out local optimum.
Using Gaussian mutation factor pair, the subgroup is disturbed.Gaussian Profile is a kind of engineering commonly important probability point
Cloth, shown in probability density function following formula:In formula, σ is the variance of Gaussian Profile, and μ is it is expected.
Gaussian mutation processing, x are carried out to the state of firefly groupi=xi+xi×N(μ,σ)。
In above-mentioned steps (6), according to the position of firefly after update, the brightness of firefly is recalculated.
After firefly may be absorbed in local optimum, Gauss disturbance is added to firefly, after Gauss disturbance is added, again
The position for updating firefly, by the new band of position people to object function of firefly, recalculating firefly brightness.
In above-mentioned steps (7), if meeting end condition, global extremum point and optimum individual value, the optimal light of firefly are exported
The correspondence database best queries executive plan of worm position, otherwise, searching times increase by 1, turn the 5th step, are searched for next time.
In above-mentioned steps 5, meet termination condition, exports best queries executive plan, searched for when using glowworm swarm algorithm
When best queries executive plan, such as reaches maximum iterations when meeting end condition, terminate search, export as a result, looking for
To best queries executive plan, the query optimization of database multi-list connection is completed.
Claims (4)
1. a kind of database multi-list Connection inquiring optimization method based on glowworm swarm algorithm, it is characterized in that:Define searching for database
Database inquiry optimization Cost Model is established in rope space;In the solution space of left deep tree composition, obtained using follow-up traversal threaded tree
To coded sequence;Define fitness function;Best queries executive plan is found using glowworm swarm algorithm;Meet termination condition, it is defeated
Go out best queries executive plan;Wherein, finding best queries executive plan algorithm steps using glowworm swarm algorithm includes:(1) just
Beginningization algorithm basic parameter, setting firefly number n, maximum Attraction Degree β0, light intensity absorption coefficient gamma, step factor α, maximum changes
Generation number MaxGeneration or search precision ε;(2) position of random initializtion firefly calculates firefly according to fitness function
The fitness value of fireworm is as its maximum fluorescence intensity I0;(3) calculate firefly relative luminance I and Attraction Degree β, relatively belonging to
The fluorescent brightness size of firefly in neighborhood determines the moving direction of firefly according to relative luminance;(4) according to the linear of introducing
Successively decrease weighting function and introduce adaptive equalization mechanism, update the position of firefly;(5) to being in the firefly of optimum position,
If optimal value does not update continuously three times, it may be absorbed in local optimum, then random perturbation is carried out to the position of optimal firefly,
The random perturbation of a Gaussian distributed is added, algorithm can be made to jump out local optimum;(6) according to the position of firefly after update
It sets, recalculates the brightness of firefly;(7) if meeting end condition, global extremum point and optimum individual value, optimal firefly are exported
The corresponding database best queries executive plan in fireworm position, otherwise, searching times increase by 1, turn (3) step, carry out next time
Search.
2. a kind of database multi-list Connection inquiring optimization method based on glowworm swarm algorithm according to claim 1, special
Sign is:Glowworm swarm algorithm is in the later stage of iteration, since firefly has slowly been moved to the attached of part or global extremum point
Closely, so the distance between firefly is gradually reduced, the Attraction Degree between firefly gradually increases, it will makes firefly individual
Displacement distance is excessive, thus can not reach or miss optimum position, causes the problem of Near The Extreme Point shakes, therefore this hair
Bright middle introducing linear decrease weighting function, weighting function formula are as follows:
Wherein, x, I points are firefly position and brightness.
3. a kind of database multi-list Connection inquiring optimization method based on glowworm swarm algorithm according to claim 1, special
Sign is:Since the firefly moving step length of algorithm design is fixed α, it is unfavorable for algorithm later stage solution local optimum, and
Adaptive step adjustment is beneficial to improve the precision and convergence rate solved, therefore present invention introduces adaptive equalization mechanism, adjust
Whole is α=α × Δ α, further increases algorithm the convergence speed and global optimizing ability.
4. a kind of database multi-list Connection inquiring optimization method based on glowworm swarm algorithm according to claim 1, special
Sign is:To being in the firefly of optimum position, if optimal value does not update continuously three times, judge that it may be absorbed in part most
It is excellent, the random perturbation of a Gaussian distributed is added to the firefly of optimal location, algorithm is made to jump out local optimum.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810216842.9A CN108388666A (en) | 2018-03-16 | 2018-03-16 | A kind of database multi-list Connection inquiring optimization method based on glowworm swarm algorithm |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810216842.9A CN108388666A (en) | 2018-03-16 | 2018-03-16 | A kind of database multi-list Connection inquiring optimization method based on glowworm swarm algorithm |
Publications (1)
Publication Number | Publication Date |
---|---|
CN108388666A true CN108388666A (en) | 2018-08-10 |
Family
ID=63067977
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810216842.9A Pending CN108388666A (en) | 2018-03-16 | 2018-03-16 | A kind of database multi-list Connection inquiring optimization method based on glowworm swarm algorithm |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108388666A (en) |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109408538A (en) * | 2018-10-22 | 2019-03-01 | 武汉达梦数据库有限公司 | Automatically cloud component is provided in cloud platform realizes extensive fusion query method and system |
CN109919294A (en) * | 2019-02-28 | 2019-06-21 | 湖北工业大学 | A kind of image enchancing method of glowworm swarm algorithm and cuckoo searching algorithm Parallel Fusion |
CN110909158A (en) * | 2019-07-05 | 2020-03-24 | 重庆信科设计有限公司 | Text classification method based on improved firefly algorithm and K nearest neighbor |
EP3627346A1 (en) * | 2018-09-20 | 2020-03-25 | Amadeus S.A.S. | Function call sequence processing |
CN111539506A (en) * | 2020-03-16 | 2020-08-14 | 中国石油天然气集团有限公司 | Firefly algorithm-based longitudinal wave time difference real-time extraction method |
CN112596373A (en) * | 2020-10-27 | 2021-04-02 | 西北工业大学 | Unmanned aerial vehicle attitude control parameter intelligent setting method based on quantum firefly algorithm |
CN113836174A (en) * | 2021-10-20 | 2021-12-24 | 昆明理工大学 | Asynchronous SQL connection query optimization method based on reinforcement learning DQN algorithm |
-
2018
- 2018-03-16 CN CN201810216842.9A patent/CN108388666A/en active Pending
Non-Patent Citations (2)
Title |
---|
刘东: "基于多子群萤火虫算法的数据库查询优化", 《计算机工程与应用》 * |
臧睿: "基于标准萤火虫算法的改进与仿真应用", 《计算机科学》 * |
Cited By (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP3627346A1 (en) * | 2018-09-20 | 2020-03-25 | Amadeus S.A.S. | Function call sequence processing |
FR3086424A1 (en) * | 2018-09-20 | 2020-03-27 | Amadeus S.A.S. | PROCESSING A SEQUENCE OF FUNCTIONAL CALLS |
US11157308B2 (en) | 2018-09-20 | 2021-10-26 | Amadeus S.A.S. | Function call sequence processing |
CN109408538A (en) * | 2018-10-22 | 2019-03-01 | 武汉达梦数据库有限公司 | Automatically cloud component is provided in cloud platform realizes extensive fusion query method and system |
CN109408538B (en) * | 2018-10-22 | 2021-06-22 | 武汉达梦数据库有限公司 | Method and system for automatically issuing cloud components in cloud platform to realize large-scale fusion query |
CN109919294A (en) * | 2019-02-28 | 2019-06-21 | 湖北工业大学 | A kind of image enchancing method of glowworm swarm algorithm and cuckoo searching algorithm Parallel Fusion |
CN110909158B (en) * | 2019-07-05 | 2022-10-18 | 重庆信科设计有限公司 | Text classification method based on improved firefly algorithm and K nearest neighbor |
CN110909158A (en) * | 2019-07-05 | 2020-03-24 | 重庆信科设计有限公司 | Text classification method based on improved firefly algorithm and K nearest neighbor |
CN111539506A (en) * | 2020-03-16 | 2020-08-14 | 中国石油天然气集团有限公司 | Firefly algorithm-based longitudinal wave time difference real-time extraction method |
CN112596373A (en) * | 2020-10-27 | 2021-04-02 | 西北工业大学 | Unmanned aerial vehicle attitude control parameter intelligent setting method based on quantum firefly algorithm |
CN112596373B (en) * | 2020-10-27 | 2023-05-23 | 西北工业大学 | Unmanned aerial vehicle attitude control parameter intelligent setting method based on quantum firefly algorithm |
CN113836174A (en) * | 2021-10-20 | 2021-12-24 | 昆明理工大学 | Asynchronous SQL connection query optimization method based on reinforcement learning DQN algorithm |
CN113836174B (en) * | 2021-10-20 | 2024-01-23 | 昆明理工大学 | Asynchronous SQL (structured query language) connection query optimization method based on reinforcement learning DQN (direct-to-inverse) algorithm |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108388666A (en) | A kind of database multi-list Connection inquiring optimization method based on glowworm swarm algorithm | |
CN110516831A (en) | A kind of short-term load forecasting method based on MWOA algorithm optimization SVM | |
Dolatshah et al. | Ball*-tree: Efficient spatial indexing for constrained nearest-neighbor search in metric spaces | |
CN101630161B (en) | Intelligent system for complex industrial production and construction method thereof | |
Tang et al. | A hybrid improved whale optimization algorithm | |
CN110070121A (en) | A kind of quick approximate k nearest neighbor method based on tree strategy with balance K mean cluster | |
CN105975587A (en) | Method for organizing and accessing memory database index with high performance | |
Bacanin et al. | RFID network planning by ABC algorithm hybridized with heuristic for initial number and locations of readers | |
CN105117461B (en) | A kind of enquiring and optimizing method based on Revised genetic algorithum | |
Ma et al. | An improved dynamic self-adaption cuckoo search algorithm based on collaboration between subpopulations | |
Cao et al. | Improved crow search algorithm optimized extreme learning machine based on classification algorithm and application | |
CN109446293A (en) | A kind of parallel higher-dimension nearest Neighbor | |
CN108615069A (en) | A kind of optimized calculation method based on improved adaptable quanta particle swarm optimization | |
Fu et al. | A novel firefly algorithm based on improved learning mechanism | |
Zhang et al. | Tent chaos and nonlinear convergence factor whale optimization algorithm | |
Du et al. | Adaptive chaos clonal evolutionary programming algorithm | |
CN111291582B (en) | RFID antenna optimal deployment method based on jellyfish group algorithm | |
Zhou et al. | An improved cuckoo search algorithm for optimizing the beam patterns of the random antenna arrays | |
CN109086941A (en) | A kind of energy-consuming prediction technique | |
CN108805254A (en) | Parameter Optimization System for time series forecasting | |
CN114004326A (en) | ELM neural network optimization method based on improved suburb algorithm | |
Liu et al. | Research on Wireless Sensor Network Localization Based on An Improved Whale Optimization Algorithm | |
Zhang et al. | K-means clustering algorithm based on bee colony strategy | |
Yin et al. | Cuckoo search algorithm based on mobile cloud model [J] | |
Krömer et al. | Differential evolution with preferential interaction network |
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: 20180810 |
|
RJ01 | Rejection of invention patent application after publication |