CN113256094B - Service resource allocation method based on improved particle swarm optimization - Google Patents

Service resource allocation method based on improved particle swarm optimization Download PDF

Info

Publication number
CN113256094B
CN113256094B CN202110533976.5A CN202110533976A CN113256094B CN 113256094 B CN113256094 B CN 113256094B CN 202110533976 A CN202110533976 A CN 202110533976A CN 113256094 B CN113256094 B CN 113256094B
Authority
CN
China
Prior art keywords
matrix
generation
formula
ith
particles
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.)
Active
Application number
CN202110533976.5A
Other languages
Chinese (zh)
Other versions
CN113256094A (en
Inventor
周子云
宋佰洋
吴悦
王子儒
易锦均
林子越
赵浩冰
崔欣
吴雨豪
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Hefei Xingbei Intelligent Control Technology Co ltd
Original Assignee
Anhui Shuaier Information Technology Co ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Anhui Shuaier Information Technology Co ltd filed Critical Anhui Shuaier Information Technology Co ltd
Priority to CN202110533976.5A priority Critical patent/CN113256094B/en
Publication of CN113256094A publication Critical patent/CN113256094A/en
Application granted granted Critical
Publication of CN113256094B publication Critical patent/CN113256094B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0631—Resource planning, allocation, distributing or scheduling for enterprises or organisations
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10—Complex mathematical operations
    • G06F17/16—Matrix or vector computation, e.g. matrix-matrix or matrix-vector multiplication, matrix factorization
    • 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]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
    • G06Q10/06393—Score-carding, benchmarking or key performance indicator [KPI] analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/04—Manufacturing
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
    • Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
    • Y02P90/30—Computing systems specially adapted for manufacturing

Landscapes

  • Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Human Resources & Organizations (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Strategic Management (AREA)
  • Economics (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Mathematical Physics (AREA)
  • General Business, Economics & Management (AREA)
  • Data Mining & Analysis (AREA)
  • Marketing (AREA)
  • Tourism & Hospitality (AREA)
  • Educational Administration (AREA)
  • Development Economics (AREA)
  • Mathematical Analysis (AREA)
  • Software Systems (AREA)
  • Mathematical Optimization (AREA)
  • Quality & Reliability (AREA)
  • Pure & Applied Mathematics (AREA)
  • General Health & Medical Sciences (AREA)
  • Computational Mathematics (AREA)
  • Health & Medical Sciences (AREA)
  • Game Theory and Decision Science (AREA)
  • General Engineering & Computer Science (AREA)
  • Computing Systems (AREA)
  • Operations Research (AREA)
  • Computational Linguistics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Biomedical Technology (AREA)
  • Biophysics (AREA)
  • Molecular Biology (AREA)
  • Evolutionary Computation (AREA)
  • Algebra (AREA)
  • Databases & Information Systems (AREA)
  • Manufacturing & Machinery (AREA)
  • Primary Health Care (AREA)
  • General Factory Administration (AREA)

Abstract

The invention discloses a service resource allocation method for improving a particle swarm algorithm, which comprises the following steps: 1, decomposing the service resource into a plurality of subtasks; 2 setting parameters of the particles; 3 generating an initial population; 4, calculating the fitness of the particles; 5, judging whether a termination condition is met, if so, outputting a global optimal solution, and if not, continuing; 6, updating the local optimal solution, the global optimal solution and the w weight of the particle; 7 updating the position and speed of the particles; and 8, adjusting the particle position, and returning to the step 4. The invention can realize the joint production of products in a plurality of factories, thereby effectively improving the utilization rate of equipment and reducing the production cost.

Description

Service resource allocation method based on improved particle swarm optimization
Technical Field
The invention relates to the field of supply chains, in particular to a service resource allocation method based on an improved particle swarm optimization, which is used for multi-factory joint production scheduling management.
Background
Today, in globalization, manufacturing enterprise supply chain collaboration is becoming more and more involved. To focus technology, resources, etc. on the key links of production, enterprises often outsource certain processes to multiple manufacturers. Due to different production capacities and different geographic positions of different manufacturers, the subtasks completed by the sub-factories in the production and transportation process have multiple indexes and high complexity.
The existing particle swarm algorithm cannot simultaneously process a plurality of complex variables in the aspect of calculating the self-adaptive degree, has no universality in the process of distributing subtask service resources on a production line, and cannot be applied in a large scale.
In a complex production task in the society, because links are multiple, workload is large, different instruments are needed for production, and multiple factories are often needed for cooperation, the problem of how to match internal links of business resources among the factories is generated, and the maximization of production efficiency is realized according to the balanced distribution of the occupancy rates of the resources of the instruments.
Disclosure of Invention
The invention aims to solve the defects in the prior art, and provides a service resource matching algorithm based on an improved particle swarm algorithm, so that different subtasks can be distributed to different factories on a production line, the cooperation efficiency among the factories can be improved, and the cooperation cost among the factories can be reduced.
In order to achieve the purpose, the invention adopts the following technical scheme:
the invention relates to a business resource allocation method based on an improved particle swarm algorithm, which is characterized in that the method is applied to the production tasks of decomposing business resources into m subtasks and allocating each subtask to n factories for processing production, and an agent alliance is formed by the n factoriesWherein, the ith subtask is marked as T i ，i∈[1，m](ii) a J th plant is marked F j ，j∈[1，n](ii) a And using binary X ij Represents the ith subtask T i And jth plant F j Thereby obtaining a service resource allocation matrix X and a service logic scheduling allocation problem matrix V:
Figure GDA0003750568500000011
in the formula (1), the reaction mixture is,
Figure GDA0003750568500000012
is the element of the ith row and the jth column in the service resource allocation matrix X, and represents the ith subtask T i And jth plant F j The positional relationship that exists; if it is
Figure GDA0003750568500000021
Represents the ith subtask T i Assigned to the i-th subtask F j Processing and producing; otherwise, the ith subtask T is represented i Not assigned to represent the ith subtask F j Processing and producing;
Figure GDA0003750568500000022
in the formula (2), v (X) ij ) Is the element of the ith row and jth column in the business logic scheduling assignment problem matrix V, which represents the ith subtask T i And jth plant F j The relationship of the increments present;
the service resource allocation method is carried out according to the following steps:
step 1, initializing each parameter in a particle swarm algorithm, comprising the following steps: iteration times L and initializing L to be 1; maximum number of iterations L max Number of two learning factors c 1 And c 2 The inertial weight ω;
step 2, making the position matrix X of the L-th generation particles L Initializing matrix X at random; speed of L-th generation particleDegree matrix V L Distributing a problem matrix V for service logic scheduling and initializing the problem matrix V to be 0 so as to generate an L-th generation population;
step 3, the L-th generation service resource distribution matrix X L The distribution situation of the medium particles obtains a two-dimensional particle matrix S:
if the L-th generation service logic scheduling distribution problem matrix X L Middle element
Figure GDA0003750568500000023
If the weight of (2) is 1, the corresponding element S in the ith row and the jth column in the two-dimensional particle matrix S shown in the formula (3) ij The weight of (1) is 1, otherwise, the element s of the ith row and the jth column ij The weight of (2) is 0;
Figure GDA0003750568500000024
in the formula (3), s ij Is to indicate the ith subtask T at the Lth generation i Assigned to the jth plant F j Generating a particle in a two-dimensional particle matrix S;
step 4, calculating the ith row and the jth column of particles S in the two-dimensional particle matrix S ij Scheduling assignment problem matrix X in Lth generation service logic L Degree of adaptability in
Figure GDA0003750568500000031
Maximum fitness of particles in the first L-1 generation
Figure GDA0003750568500000032
Degree of adaptability
Figure GDA0003750568500000033
Make a comparison if
Figure GDA0003750568500000034
Then the maximum fitness will be
Figure GDA0003750568500000035
Corresponding element
Figure GDA0003750568500000036
Is used as the Lth generation service logic scheduling distribution problem matrix X L The best position of the ith row and the jth column of particles in the search space
Figure GDA0003750568500000037
Otherwise, the fitness is measured
Figure GDA0003750568500000038
Corresponding element
Figure GDA0003750568500000039
Is used as the L-th generation service logic scheduling distribution problem matrix X L The best position of the ith row and the jth column of particles in the search space
Figure GDA00037505685000000310
Thereby obtaining the local optimal solution matrix of the particles in the two-dimensional particle matrix S at the L-th generation
Figure GDA00037505685000000311
Figure GDA00037505685000000312
Step 5, from local optimal solution matrix of L generation
Figure GDA00037505685000000313
The position of the optimal solution corresponding to the maximum fitness of the ith row is selected as a position matrix X L The global optimal solution of the ith row is recorded as
Figure GDA00037505685000000314
Otherwise, returning to the step 4 for sequential execution; wherein,
Figure GDA00037505685000000315
representing a position matrix X L The global optimal solution of the ith row;
step 6, solving a velocity matrix V of the particles in the L generation L ；
Step 6.1, calculating an inertia factor omega by using the formula (5):
ω＝(ω ini -ω end )(L max -L)/L max +ω end (5)
in the formula (5), L max Is the maximum number of iterations, L is the current number of iterations, ω ini As an initial inertia weight, ω end The inertia weight value is the inertia weight value when iteration is carried out to the maximum evolution algebra;
step 6.2, updating the velocity matrix V of the L-th generation particles by the formula (6) L ：
Figure GDA00037505685000000316
In the formula (6), c 1 ,c 2 Is a learning factor; rand is [0, 1 ]]A uniform random number within a range; when L is 1, V L-1 ＝0；
Step 7, calculating the L +1 th generation position matrix X according to the formula (7) L+1 ：
X L+1 ＝X L +V L (7)
Step 8, calculating the fitness of the jth particle in the ith row of the L +1 generation in the two-dimensional particle matrix S
Figure GDA00037505685000000317
And the fitness of the corresponding particles in the L generation
Figure GDA0003750568500000041
Comparing, and taking the particle position corresponding to the larger fitness value as the optimal solution position of the jth particle in the ith row of the L +1 th generation in the two-dimensional particle matrix S
Figure GDA0003750568500000042
Step 9, repeating step 8 to calculate the fitness of the next particle of the L +1 th generation in the two-dimensional particle matrix S and comparing the fitness to obtain the L + of the two-dimensional particle matrix SOptimal solution matrix of 1 generation all particles
Figure GDA0003750568500000043
From the optimal solution matrix
Figure GDA0003750568500000044
The optimal solution corresponding to the maximum fitness is selected from the ith row of the position matrix X as the L +1 th generation L+1 Global optimal solution for row i of
Figure GDA0003750568500000045
Step 10, assigning L +1 to L, and judging L<L max If yes, returning to the step 4 for sequential execution, otherwise, indicating that L is finished max The second iteration to obtain the L max Surrogate location matrix X Lmax Global optimal solution for row i of
Figure GDA0003750568500000046
Thereby obtaining a two-dimensional matrix
Figure GDA00037505685000000414
Global optimal solution for each row
Figure GDA0003750568500000047
Namely, it is
Figure GDA0003750568500000048
At the L th max Position moment matrix
Figure GDA0003750568500000049
The upper position indicates the optimal plant to which the i-th sub-task is assigned.
The service resource allocation method based on the improved particle swarm optimization is also characterized in that the fitness of the particles in the step 4
Figure GDA00037505685000000410
The calculation is carried out according to the following steps:
step 4.1, dividing the evaluation indexes into accurate indexes and grade indexes;
4.2, converting the low-quality index in the accurate index into a high-quality index by using the formula (8), thereby obtaining a factory evaluation index matrix H shown as the formula (9);
Figure GDA00037505685000000411
Figure GDA00037505685000000412
in the formula (9), h ij A j high-quality index representing the ith factory;
Figure GDA00037505685000000413
a j-th low-priority index representing an i-th plant;
4.3, carrying out normalization processing on the factory evaluation index matrix H to obtain an index homotaxial normalization matrix Z shown in a formula (10);
Figure GDA0003750568500000051
in the formula (10), Z ij Representing the jth index value of the ith plant in the index homotaxial normalization matrix;
step 4.4, according to the index homotaxial normalization matrix Z, respectively determining the optimal scheme Z by using the formula (11) and the formula (12) + And the worst case Z - ：
Figure GDA0003750568500000052
Figure GDA0003750568500000053
In the formulae (11) and (12),
Figure GDA0003750568500000054
the maximum value of the ith column index in the index homotaxial normalization matrix Z is represented;
Figure GDA0003750568500000055
the minimum value of the ith column index in the index homotaxial normalization matrix Z is represented;
step 4.5, solving the current scheme and the optimal scheme Z by using the formula (13) and the formula (14) + Is a distance of
Figure GDA0003750568500000056
With the current scheme and the worst scheme
Figure GDA0003750568500000057
Is a distance of
Figure GDA0003750568500000058
Figure GDA0003750568500000059
Figure GDA00037505685000000510
Step 4.6, calculating the self-adaptability of the ith task distributed to j plant schemes by using the formula (15)
Figure GDA00037505685000000511
Figure GDA00037505685000000512
Compared with the prior art, the invention has the beneficial effects that:
1. according to the invention, under a typical differential batch manufacturing mode, the problem of cooperative scheduling production of enterprises is researched, and by adopting an improved particle swarm algorithm, the problems that evaluation indexes are not uniform in the production and transportation process, multiple indexes need to be considered during distribution, indexes of subtasks completed by a factory are unified, and the method has universality and adaptability. Obtaining the position and the speed of the particles; then, updating the positions and the speeds of the particles by utilizing a particle swarm algorithm, realizing multiple iterations and finally obtaining an optimal solution;
2. the invention provides a novel service resource allocation method based on a particle swarm algorithm. Firstly, complex business is decomposed into a plurality of flow segments, and then the optimal solution is worked out by dynamic planning of a particle swarm algorithm. The method can effectively solve the problem of current complex business resource allocation (instrument and service allocation), and has accuracy, rapidity and robustness. The method has strong universality when calculating the self-adaptability, can be widely applied and solves the problems of distribution and scheduling of various service resources.
3. The invention decomposes the complex task into a plurality of normalized subtask flows according to the business logic relationship, forms an Agent production union among the small factories to form a complete production chain, decomposes the complex production task into different subtasks with logic relationship when the user demand is generated, disperses the different subtasks to each small factory based on the particle swarm algorithm, and selects a proper path from the different subtasks. The algorithm has strong universality, can unify complex indexes in the production and transportation processes, has strong universality, and can be applied to the field of different business logic distribution.
Drawings
FIG. 1 is a flow chart of a method for allocating service resources by using a particle swarm algorithm according to the present invention;
FIG. 2 is a flow chart of particle fitness calculation according to the present invention.
Detailed Description
In this embodiment, a service resource allocation method for improving a particle swarm algorithm is to divide a service resource into m subtasks, allocate each subtask to n production tasks for processing and production in n factories, and form an agent union by the n factories, where the ith subtask is denoted as T i ，i∈[1，m](ii) a J th plant is marked F j ，j∈[1，n](ii) a And using binary X ij Represents the ith subtask T i And jth plant F j Thereby obtaining a service resource allocation matrix X and a service logic scheduling allocation problem matrix V:
Figure GDA0003750568500000061
in the formula (1), the acid-base catalyst,
Figure GDA0003750568500000062
is the element of the ith row and jth column in the service resource allocation matrix X, which represents the ith subtask T i And jth plant F j The positional relationship that exists; if it is
Figure GDA0003750568500000063
Represents the ith subtask T i Assigned to the i-th subtask F j Processing and producing; otherwise, the ith subtask T is represented i Not assigned to represent the ith subtask F j Processing and producing;
Figure GDA0003750568500000071
in the formula (2), v (X) ij ) Is the element of the ith row and jth column in the business logic scheduling assignment problem matrix V, which represents the ith subtask T i And the jth plant F j The relationship of the increments present;
the service resource allocation method is performed according to the following steps as shown in fig. 1:
step 1, initializing each parameter in a particle swarm algorithm, comprising the following steps: iteration times L and initializing L to 1; maximum number of iterations L max Number of two learning factors c 1 And c 2 The inertial weight ω, given by L max Has a value of 300, learning factor c 1 And c 2 Has a value of 2 and the inertial weight ω has a value of 1;
step 2, making the position matrix X of the L-th generation particles L For matrix X and randomly initializing to generate L generation population, and distributing v (X) in problem matrix according to sum service logic scheduling ij ) Obtaining the value of the service resource distribution matrix X L Initial position and initial velocity of the ith row and jth column of the particle, if
Figure GDA0003750568500000072
If the value of (1) is less than the threshold, it represents the service resource allocation matrix X L The ith row and the jth column of the medium-speed particle matrix have the existence of particles, otherwise, the No is not, so that the speed matrix V of the L-th generation particles L Assigning problem matrix for service logic scheduling order service logic scheduling assignment problem matrix V L V (X) in ij ) Initialization to 0, where v (X) ij ) The particle is represented in the service resource allocation matrix X L The velocity increment of the ith row and the jth column of particles;
step 3, the L-th generation service resource distribution matrix X L Obtaining a two-dimensional particle matrix S according to the distribution condition of the medium particles:
if the Lth generation service logic scheduling distribution problem matrix X L Middle element
Figure GDA0003750568500000073
If the weight of (2) is 1, the corresponding element S in the ith row and the jth column in the two-dimensional particle matrix S shown in the formula (3) ij The weight of (1) is 1, otherwise, the element s of the ith row and the jth column ij The weight of (2) is 0;
Figure GDA0003750568500000074
in the formula (3), s ij Is to indicate the ith subtask T at the Lth generation i Assigned to the jth plant F j Generating a particle in a two-dimensional particle matrix S, each particle having memory, i.e. when the number of iterations is L-th generation ij The fitness of L-1 generation particles containing L generation particles can be recorded
Figure GDA0003750568500000081
Step 4, calculating the ith row and the jth column of particles S in the two-dimensional particle matrix S ij Scheduling assignment problem matrix X in Lth generation service logic L Degree of adaptability in
Figure GDA0003750568500000082
Maximum fitness of particles in the first L-1 generation
Figure GDA0003750568500000083
Degree of adaptability
Figure GDA0003750568500000084
Make a comparison if
Figure GDA0003750568500000085
Then the maximum fitness will be
Figure GDA0003750568500000086
Corresponding element
Figure GDA0003750568500000087
Is used as the L-th generation service logic scheduling distribution problem matrix X L The best position of the ith row and jth column of particles in the search space
Figure GDA0003750568500000088
Otherwise, the fitness is measured
Figure GDA0003750568500000089
Corresponding element
Figure GDA00037505685000000810
Is used as the Lth generation service logic scheduling distribution problem matrix X L The best position of the ith row and jth column of particles in the search space
Figure GDA00037505685000000811
Thereby obtaining the local optimal solution matrix of the particles in the two-dimensional particle matrix S at the L-th generation
Figure GDA00037505685000000812
Figure GDA00037505685000000813
Fitness of the particles in step 4
Figure GDA00037505685000000814
The calculation was performed as follows as shown in fig. 2:
and 4.1, recording indexes in the production and transportation process of a factory, and establishing an evaluation index model. The evaluation indexes are divided into accurate indexes (objective indexes) and grade indexes (subjective indexes) according to the factory reference indexes in the actual ordering process, and the model is used for calculating and updating the self-adaptability of the factory. The method mainly comprises two parts, namely an accurate index and a grade index, wherein the accurate index comprises transportation cost, transportation time, production time, product price and the like; the grade index comprises a factory quality score, a service attitude and the like.
According to the consideration factors of the actual ordering process, the model uses the accurate indexes of transportation cost, transportation time, production time, product price, factory quality score and grade index of service attitude as shown in table 1.
TABLE 1 index Classification
Figure GDA00037505685000000815
Taking the selected indexes as the reference of the selected factory, defining the effective interval of each index as the following table 2, wherein the effective interval of the quality score and the service attitude of the factory is { A, B, C, D, E }, and the effective interval of the transportation cost, the transportation time, the production time and the product price is [0, ∞ ].
TABLE 2 evaluation index intervals for factories
Figure GDA0003750568500000091
The two indexes are assigned in respective intervals by using different assignment methods, and the first accurate index is assigned according to an actual value; the second class index is assigned according to the class, a being a value of 5, B being a value of 4, C being a value of 3, D being a value of 2, and E being a value of 1.
Step 4.2, performing chemotaxis treatment on the evaluation indexes, and converting low-quality indexes in the accurate indexes into high-quality indexes by using the formula (1), so as to obtain a factory evaluation index matrix H shown in the formula (2);
Figure GDA0003750568500000092
Figure GDA0003750568500000093
in the formula (2), h ij A j high-quality index representing the ith factory;
Figure GDA0003750568500000094
a j-th low-priority index representing an i-th plant;
4.3, carrying out normalization processing on the indexes of each factory in the factory evaluation index matrix H by using the formula (3) to obtain an index homotaxial normalization matrix Z shown as the formula (4), wherein Z ij Representing the jth index value of the ith plant in the index homotaxial normalization matrix;
Figure GDA0003750568500000101
Figure GDA0003750568500000102
and 4.4, performing homotaxial normalization processing on the indexes, having universality, making decisions by taking various indexes as judgment conditions at the same time, and respectively determining an optimal scheme Z consisting of optimal evaluation indexes in current production and transportation evaluation by using a formula (5) and a formula (6) according to an index homotaxial normalization matrix Z + And the worst scheme Z consisting of the worst evaluation index - ：
Figure GDA0003750568500000103
Figure GDA0003750568500000104
In the formulae (5) and (6),
Figure GDA0003750568500000105
the maximum value of the ith column index in the index homotaxial normalization matrix Z is represented;
Figure GDA0003750568500000106
the minimum value of the ith column index in the index homotaxial normalization matrix Z is represented;
step 4.5, solving the current scheme and the optimal scheme Z by using the formula (7) and the formula (8) + Is a distance of
Figure GDA0003750568500000107
With the current scheme and the worst scheme
Figure GDA0003750568500000108
Is a distance of
Figure GDA0003750568500000109
Figure GDA00037505685000001010
Figure GDA00037505685000001011
Step 4.6, calculating the self-adaptability of the ith task distributed to j plant schemes by using the formula (9)
Figure GDA00037505685000001012
Figure GDA00037505685000001013
According to
Figure GDA00037505685000001014
To the optimal solution Z + If Z is ij For the optimal solution, the corresponding T i 1 is ═ 1; if Z is ij Is the worst scheme, then corresponding T i 0. If it is
Figure GDA00037505685000001015
The closer to 1, Z j The closer to the optimal solution; if it is
Figure GDA00037505685000001016
The closer to 0, Z ij The closer to the worst case to derive the degree of adaptability of each plant.
Step 5, from local optimal solution matrix of L generation
Figure GDA0003750568500000111
The position of the optimal solution corresponding to the maximum fitness of the ith row is selected as a position matrix X L The global optimal solution of the ith row is recorded as
Figure GDA0003750568500000112
Otherwise, returning to the step 4 for sequential execution; wherein,
Figure GDA0003750568500000113
representing a position matrix X L The global optimal solution of the ith row;
step 6.1, the service resource allocation algorithm based on the improved particle swarm optimization is shown in fig. 1, and when the iteration times are L generations, an inertia factor ω is calculated by using a formula (10):
ω＝(ω ini -ω end )(L max -L)/L max +ω end (10)
in the formula (10), L max Is the maximum number of iterations, ω ini Is an initial inertia weight, omega end The inertia weight when iterating to the maximum evolution algebra.
Step 6.2, updating the velocity matrix V of the L-th generation particles by the formula (11) L ：
Figure GDA0003750568500000114
In the formula (11), c 1 ,c 2 Is a learning factor; rand () is [0, 1 ]]The uniform random number in the range, the formula of the update speed is composed of three parts, the first part is omega V L-1 The 'inertia' reflects the 'habit' of the movement of the subtask particle swarm; the second part
Figure GDA0003750568500000115
Is 'cognitive', reflects memory of self historical experience (i.e. best-in-proximity to self); the third part
Figure GDA0003750568500000116
Is 'society', and the group historical experience reflecting the cooperative knowledge sharing among the subtask particles approaches to the global optimum.
Step 7, calculating the L +1 th generation position matrix X according to the formula (12) L+1 ：
X L+1 ＝X L +V L (12)
Step 8, calculating the fitness of the jth particle in the ith row of the L +1 generation in the two-dimensional particle matrix S
Figure GDA0003750568500000117
And the fitness of the corresponding particles in the L generation
Figure GDA0003750568500000118
Comparing, and taking the particle position corresponding to the larger fitness value as the optimal solution position of the jth particle in the ith row of the L +1 th generation in the two-dimensional particle matrix S
Figure GDA0003750568500000119
Step (ii) of9. Repeating the step 8 to calculate the fitness of the next particle in the L +1 th generation in the two-dimensional particle matrix S and comparing the fitness to obtain the optimal solution matrix of all the particles in the L +1 th generation in the two-dimensional particle matrix S
Figure GDA00037505685000001110
From the optimal solution matrix
Figure GDA00037505685000001111
The optimal solution corresponding to the maximum fitness is selected from the ith row of the matrix as the L + 1-th generation position matrix X L+1 Global optimal solution for ith row
Figure GDA0003750568500000121
Step 10, assigning L +1 to L, and judging L<L max If yes, returning to the step 4 for sequential execution, otherwise, indicating that L is finished max The second iteration to obtain the L max Surrogate position matrix X Lmax Global optimal solution for row i of
Figure GDA0003750568500000122
Thereby obtaining a two-dimensional matrix
Figure GDA0003750568500000123
Global optimal solution for each row
Figure GDA0003750568500000124
Namely that
Figure GDA0003750568500000125
At the L th max Position moment matrix
Figure GDA0003750568500000126
The upper position indicates the optimal plant to which the i-th sub-task is assigned.

Claims (2)

1. A service resource allocation method based on improved particle swarm optimization is characterized in that the method is applied to decomposing service resources into m subtasks and allocating each subtask toIn the production tasks of processing production of n factories, an agent alliance is formed by the n factories, wherein the ith subtask is recorded as T i ，i∈[1，m](ii) a J th plant is marked F j ，j∈[1，n](ii) a And using binary X ij Represents the ith subtask T i And the jth plant F j Thereby obtaining a service resource allocation matrix X and a service logic scheduling allocation problem matrix V:
Figure FDA0003750568490000011
in the formula (1), the acid-base catalyst,
Figure FDA0003750568490000012
is the element of the ith row and the jth column in the service resource allocation matrix X, and represents the ith subtask T i And the jth plant F j The positional relationship that exists; if it is
Figure FDA0003750568490000013
Represents the ith subtask T i Assigned to the i-th subtask F j Processing and producing; otherwise, the ith subtask T is represented i Not assigned to represent the ith subtask F j Processing and producing;
Figure FDA0003750568490000014
in the formula (2), v (X) ij ) Is the element of ith row and jth column in the service logic scheduling assignment problem matrix V, which represents the ith subtask T i And jth plant F j The relationship of the increments present;
the service resource allocation method is carried out according to the following steps:
step 1, initializing each parameter in a particle swarm algorithm, comprising the following steps: iteration times L and initializing L to be 1; maximum number of iterations L max Number of two learning factors c 1 And c 2 The inertial weight ω;
step 2, enabling the position matrix X of the L-th generation particles L Initializing matrix X at random; let the velocity matrix V of the L-th generation particle L Distributing a problem matrix V for service logic scheduling and initializing the problem matrix V to be 0 so as to generate an L-th generation population;
step 3, the L-th generation service resource distribution matrix X L The distribution situation of the medium particles obtains a two-dimensional particle matrix S:
if the L-th generation service logic scheduling distribution problem matrix X L Middle element
Figure FDA0003750568490000015
If the weight of (2) is 1, the corresponding element S in the ith row and the jth column in the two-dimensional particle matrix S shown in the formula (3) ij The weight of (1) is 1, otherwise, the element s of the ith row and the jth column ij The weight of (2) is 0;
Figure FDA0003750568490000021
in the formula (3), s ij Is to indicate the ith subtask T in the Lth generation i Assigned to the jth plant F j Generating a particle in a two-dimensional particle matrix S;
step 4, calculating the ith row and the jth column of particles S in the two-dimensional particle matrix S ij Scheduling assignment problem matrix X in Lth generation service logic L Degree of adaptability in
Figure FDA0003750568490000022
Maximum fitness of particles in the first L-1 generation
Figure FDA0003750568490000023
Degree of adaptability
Figure FDA0003750568490000024
Make a comparison if
Figure FDA0003750568490000025
Then the maximum fitness will be
Figure FDA0003750568490000026
Corresponding element
Figure FDA0003750568490000027
Is used as the L-th generation service logic scheduling distribution problem matrix X L The best position of the ith row and jth column of particles in the search space
Figure FDA0003750568490000028
Otherwise, the fitness is measured
Figure FDA0003750568490000029
Corresponding element
Figure FDA00037505684900000210
Is used as the Lth generation service logic scheduling distribution problem matrix X L The best position of the ith row and the jth column of particles in the search space
Figure FDA00037505684900000211
Thereby obtaining the local optimal solution matrix of the particles in the two-dimensional particle matrix S at the L-th generation
Figure FDA00037505684900000212
Figure FDA00037505684900000213
Step 5, from local optimal solution matrix of L generation
Figure FDA00037505684900000214
The position of the optimal solution corresponding to the maximum fitness of the ith row is selected as a position matrix X L The global optimal solution of the ith row is recorded as
Figure FDA00037505684900000215
Otherwise, returning to the step 4 for sequential execution; wherein,
Figure FDA00037505684900000216
representing a position matrix X L The global optimal solution of the ith row;
step 6, solving a velocity matrix V of the particles in the L generation L ；
Step 6.1, calculating an inertia factor omega by using the formula (5):
ω＝(ω ini -ω end )(L max -L)/L max +ω end (5)
in the formula (5), L max Is the maximum number of iterations, L is the current number of iterations, ω ini Is an initial inertia weight, omega end The inertia weight value is the inertia weight value when iteration is carried out to the maximum evolution algebra;
step 6.2, updating the velocity matrix V of the L-th generation particles by the formula (6) L ：
Figure FDA0003750568490000031
In the formula (6), c 1 ,c 2 Is a learning factor; rand is [0, 1 ]]A uniform random number within a range; when L is 1, V L-1 ＝0；
Step 7, calculating the L +1 th generation position matrix X according to the formula (7) L+1 ：
X L+1 ＝X L +V L (7)
Step 8, calculating the fitness of the jth particle in the ith row of the L +1 generation in the two-dimensional particle matrix S
Figure FDA0003750568490000032
And the fitness of the corresponding particles in the L generation
Figure FDA0003750568490000033
Comparing the positions of the particles with larger fitness value as twoOptimal solution position of ith row and jth particle in L +1 generation in dimension particle matrix S
Figure FDA0003750568490000034
Step 9, repeating step 8 to calculate the fitness of the next particle of the L +1 th generation in the two-dimensional particle matrix S and compare the fitness to obtain the optimal solution matrix of all the particles of the L +1 th generation in the two-dimensional particle matrix S
Figure FDA0003750568490000035
From the optimal solution matrix
Figure FDA0003750568490000036
The optimal solution corresponding to the maximum fitness is selected from the ith row of the matrix as the L + 1-th generation position matrix X L+1 Global optimal solution for row i of
Figure FDA0003750568490000037
Step 10, assigning L +1 to L, and judging L<L max If yes, returning to the step 4 for sequential execution, otherwise, indicating that L is finished max Performing secondary iteration to obtain L max Surrogate position matrix X Lmax Global optimal solution for row i of
Figure FDA0003750568490000038
Thereby obtaining a two-dimensional matrix X Lmax Global optimal solution per line
Figure FDA0003750568490000039
Namely, it is
Figure FDA00037505684900000310
At the L th max Position moment matrix
Figure FDA00037505684900000311
The upper position indicates the optimal plant assigned by the i-th sub-task.
2. The improved PSO-based service resource allocation method as claimed in claim 1, wherein the fitness of the particles in step 4 is
Figure FDA00037505684900000312
The calculation is carried out according to the following steps:
step 4.1, dividing the evaluation indexes into accurate indexes and grade indexes;
4.2, converting the low-quality index in the accurate index into a high-quality index by using the formula (8), thereby obtaining a factory evaluation index matrix H shown as the formula (9);
Figure FDA0003750568490000041
Figure FDA0003750568490000042
in the formula (9), h ij A j high-quality index representing the ith plant;
Figure FDA0003750568490000043
a j-th low-priority index representing an i-th plant;
4.3, carrying out normalization processing on the factory evaluation index matrix H to obtain an index homodyne normalization matrix Z shown as the formula (10);
Figure FDA0003750568490000044
in the formula (10), Z ij Expressing the jth index value of the ith factory in the homotaxial normalization matrix of the index;
step 4.4, according to the index homotaxial normalization matrix Z, respectively determining the optimal scheme Z by using the formula (11) and the formula (12) + And worst case Z - ：
Figure FDA0003750568490000045
Figure FDA0003750568490000046
In the formulae (11) and (12),
Figure FDA0003750568490000047
the maximum value of the ith column index in the index homotaxial normalization matrix Z is represented;
Figure FDA0003750568490000048
the minimum value of the ith column index in the index homotaxial normalization matrix Z is represented;
step 4.5, solving the current scheme and the optimal scheme Z by using the formula (13) and the formula (14) + Is a distance of
Figure FDA0003750568490000049
With the current scheme and the worst scheme
Figure FDA00037505684900000410
Is a distance of
Figure FDA00037505684900000411
Figure FDA00037505684900000412
Figure FDA0003750568490000051
Step 4.6, calculating the self-adaptability of the ith task distributed to j plant schemes by using the formula (15)
Figure FDA0003750568490000052
Figure FDA0003750568490000053
CN202110533976.5A 2021-05-17 2021-05-17 Service resource allocation method based on improved particle swarm optimization Active CN113256094B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110533976.5A CN113256094B (en) 2021-05-17 2021-05-17 Service resource allocation method based on improved particle swarm optimization

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110533976.5A CN113256094B (en) 2021-05-17 2021-05-17 Service resource allocation method based on improved particle swarm optimization

Publications (2)

Publication Number Publication Date
CN113256094A CN113256094A (en) 2021-08-13
CN113256094B true CN113256094B (en) 2022-09-13

Family

ID=77183383

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110533976.5A Active CN113256094B (en) 2021-05-17 2021-05-17 Service resource allocation method based on improved particle swarm optimization

Country Status (1)

Country Link
CN (1) CN113256094B (en)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN118691331B (en) * 2024-08-27 2024-11-01 浙江湖州开芯数字科技有限公司 Digital enterprise service resource management system and method based on industry support

Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104536828A (en) * 2014-12-26 2015-04-22 湖南强智科技发展有限公司 Cloud computing task scheduling method and system based on quantum-behaved particle swarm algorithm
CN105069010A (en) * 2015-07-07 2015-11-18 西安电子科技大学 Resource polymerization method based on Agent
CN105956689A (en) * 2016-04-21 2016-09-21 合肥工业大学 Transportation and production coordinated scheduling method based on improved particle swarm optimization
CN108320080A (en) * 2018-01-05 2018-07-24 上海电力学院 The real-time dynamic power allocation method in energy internet based on two layers of consistency algorithm
CN109214695A (en) * 2018-09-18 2019-01-15 合肥工业大学 Based on the high-end equipment research and development manufacture coordinated dispatching method and system for improving EDA
CN109670689A (en) * 2018-12-06 2019-04-23 杭州电子科技大学 A kind of scientific workflow Multiobjective Scheduling method based on immunity particle cluster
CN109687532A (en) * 2019-03-08 2019-04-26 燕山大学 A kind of combined heat and power dispatching method improving wind electricity digestion based on cooperative game
CN110363402A (en) * 2019-06-26 2019-10-22 同济大学 A kind of factory personnel dispatching method based on grouping strategy
CN110533301A (en) * 2019-08-09 2019-12-03 大连理工大学 A kind of population dispatching method based on dynamic constrained matrix

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8315901B2 (en) * 2007-05-30 2012-11-20 Verint Systems Inc. Systems and methods of automatically scheduling a workforce
CN104392298A (en) * 2014-10-28 2015-03-04 沈阳建筑大学 Production scheduling method adopting particle swarm algorithm
US20170116522A1 (en) * 2015-10-05 2017-04-27 Telekom Malaysia Berhad Method For Task Scheduling And Resources Allocation And System Thereof
CN107203412A (en) * 2017-05-04 2017-09-26 电子科技大学 A kind of cloud resource method for optimizing scheduling that particle cluster algorithm is improved based on membranous system

Patent Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104536828A (en) * 2014-12-26 2015-04-22 湖南强智科技发展有限公司 Cloud computing task scheduling method and system based on quantum-behaved particle swarm algorithm
CN105069010A (en) * 2015-07-07 2015-11-18 西安电子科技大学 Resource polymerization method based on Agent
CN105956689A (en) * 2016-04-21 2016-09-21 合肥工业大学 Transportation and production coordinated scheduling method based on improved particle swarm optimization
CN108320080A (en) * 2018-01-05 2018-07-24 上海电力学院 The real-time dynamic power allocation method in energy internet based on two layers of consistency algorithm
CN109214695A (en) * 2018-09-18 2019-01-15 合肥工业大学 Based on the high-end equipment research and development manufacture coordinated dispatching method and system for improving EDA
CN109670689A (en) * 2018-12-06 2019-04-23 杭州电子科技大学 A kind of scientific workflow Multiobjective Scheduling method based on immunity particle cluster
CN109687532A (en) * 2019-03-08 2019-04-26 燕山大学 A kind of combined heat and power dispatching method improving wind electricity digestion based on cooperative game
CN110363402A (en) * 2019-06-26 2019-10-22 同济大学 A kind of factory personnel dispatching method based on grouping strategy
CN110533301A (en) * 2019-08-09 2019-12-03 大连理工大学 A kind of population dispatching method based on dynamic constrained matrix

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
Ant Colony Algorithm for Job Scheduling in Grid Computing;Ku Ruhana Ku-Mahamud;《IEEE》;20100621;全文 *

Also Published As

Publication number Publication date
CN113256094A (en) 2021-08-13

Similar Documents

Publication Publication Date Title
Weckenborg et al. Balancing of assembly lines with collaborative robots
Hamzadayi et al. Modeling and solving static m identical parallel machines scheduling problem with a common server and sequence dependent setup times
Kumanan et al. Multi-project scheduling using an heuristic and a genetic algorithm
Chen A hybrid algorithm for allocating tasks, operators, and workstations in multi-manned assembly lines
Nearchou Balancing large assembly lines by a new heuristic based on differential evolution method
Che A multi-objective optimization algorithm for solving the supplier selection problem with assembly sequence planning and assembly line balancing
CN114186791A (en) Multi-model small-batch-oriented dynamic scheduling method for debugging production of complex equipment products
Huang et al. Genetic algorithm for rotary machine scheduling with dependent processing times
CN113256094B (en) Service resource allocation method based on improved particle swarm optimization
CN110928261A (en) Distributed estimation scheduling method and system for distributed heterogeneous flow shop
CN105760621A (en) Assembly line balance method drawing complexity into consideration
CN117952009A (en) Intelligent production line testable digital twin modeling method
CN111210125B (en) Multi-target workpiece batch scheduling method and device based on historical information guidance
Cao et al. A bottleneck degree-based migrating birds optimization algorithm for the PCB production scheduling
Yang et al. Mathematical modeling of multi-plant order allocation problem and solving by genetic algorithm with matrix representation
Liu et al. Integrated optimization of mixed-model assembly line balancing and buffer allocation based on operation time complexity
CN106611381A (en) Algorithm for analyzing influence of material purchase to production scheduling of manufacturing shop based on cloud manufacturing
Ferretti et al. Q-Learning for Inventory Management: an application case
Chaudhuri et al. A goal programming embedded genetic algorithm for multi-objective manufacturing cell design
CN117808246A (en) Flexible job shop scheduling method, device and system
D'Amico et al. Lot streaming Permutation Flow shop with energy awareness
Ezugwu Metaheuristic Optimization for Sustainable Unrelated Parallel Machine Scheduling: A concise overview with a proof-of-concept study
Chiu et al. Improving the efficiency of evolutionary algorithms for large-scale optimization with multi-fidelity models
Chaudhuri et al. A review on the multi-objective cell formation problem in cellular manufacturing systems
Ehtesham Rasi Optimization of the multi-objective flexible job shop scheduling model by applying NSGAII and NRGA algorithms

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
GR01 Patent grant
GR01 Patent grant
TR01 Transfer of patent right

Effective date of registration: 20240102

Address after: No. 369 Huayuan Avenue, Baohe Economic Development Zone, Hefei City, Anhui Province, 230041, F326

Patentee after: HEFEI SSTARS MONITORING INFORMATION TECHNOLOGY Co.,Ltd.

Address before: Room c503, feicui science and education building, Hefei University of technology, 485 Danxia Road, University Town, Hefei Economic and Technological Development Zone, 230000, Anhui Province

Patentee before: Anhui shuaier Information Technology Co.,Ltd.

TR01 Transfer of patent right
TR01 Transfer of patent right

Effective date of registration: 20240604

Address after: Room F320, Intelligent Manufacturing Technology Research Institute, Hefei University of Technology, No. 369 Huayuan Avenue, Baohe District, Hefei City, Anhui Province, 230051

Patentee after: Hefei Xingbei Intelligent Control Technology Co.,Ltd.

Country or region after: China

Address before: No. 369 Huayuan Avenue, Baohe Economic Development Zone, Hefei City, Anhui Province, 230041, F326

Patentee before: HEFEI SSTARS MONITORING INFORMATION TECHNOLOGY Co.,Ltd.

Country or region before: China

TR01 Transfer of patent right