CN117217487A - Resource scheduling method based on improved multi-neighborhood local enhancement search algorithm - Google Patents
Resource scheduling method based on improved multi-neighborhood local enhancement search algorithm Download PDFInfo
- Publication number
- CN117217487A CN117217487A CN202311274018.6A CN202311274018A CN117217487A CN 117217487 A CN117217487 A CN 117217487A CN 202311274018 A CN202311274018 A CN 202311274018A CN 117217487 A CN117217487 A CN 117217487A
- Authority
- CN
- China
- Prior art keywords
- population
- local
- individual
- representing
- resource scheduling
- 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
- 238000000034 method Methods 0.000 title claims abstract description 85
- 238000010845 search algorithm Methods 0.000 title claims abstract description 20
- 239000011159 matrix material Substances 0.000 claims abstract description 33
- 238000011423 initialization method Methods 0.000 claims abstract description 8
- 238000013507 mapping Methods 0.000 claims abstract description 7
- 230000005012 migration Effects 0.000 claims description 32
- 238000013508 migration Methods 0.000 claims description 32
- 230000006870 function Effects 0.000 claims description 30
- 230000008569 process Effects 0.000 claims description 29
- 229940088597 hormone Drugs 0.000 claims description 24
- 239000005556 hormone Substances 0.000 claims description 24
- 238000000605 extraction Methods 0.000 claims description 21
- 238000005265 energy consumption Methods 0.000 claims description 15
- 238000012545 processing Methods 0.000 claims description 14
- 239000002245 particle Substances 0.000 claims description 11
- 230000006399 behavior Effects 0.000 claims description 10
- 230000007246 mechanism Effects 0.000 claims description 8
- 230000009123 feedback regulation Effects 0.000 claims description 7
- 230000009326 social learning Effects 0.000 claims description 7
- 230000003542 behavioural effect Effects 0.000 claims description 6
- 238000003754 machining Methods 0.000 claims description 6
- 230000004044 response Effects 0.000 claims description 6
- 125000004122 cyclic group Chemical group 0.000 claims description 3
- 230000009977 dual effect Effects 0.000 claims description 3
- 238000004519 manufacturing process Methods 0.000 abstract description 22
- 238000004422 calculation algorithm Methods 0.000 abstract description 15
- 238000010586 diagram Methods 0.000 description 8
- 238000004590 computer program Methods 0.000 description 7
- 210000000349 chromosome Anatomy 0.000 description 5
- 238000005457 optimization Methods 0.000 description 5
- 230000002028 premature Effects 0.000 description 4
- 230000028327 secretion Effects 0.000 description 4
- 238000006243 chemical reaction Methods 0.000 description 3
- 238000013468 resource allocation Methods 0.000 description 3
- 238000003860 storage Methods 0.000 description 3
- 230000006978 adaptation Effects 0.000 description 2
- 230000001174 ascending effect Effects 0.000 description 2
- 230000003247 decreasing effect Effects 0.000 description 2
- 230000013016 learning Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000013386 optimize process Methods 0.000 description 2
- 108090000623 proteins and genes Proteins 0.000 description 2
- 238000012163 sequencing technique Methods 0.000 description 2
- 238000012893 Hill function Methods 0.000 description 1
- 238000003491 array Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 210000004027 cell Anatomy 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000012885 constant function Methods 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 244000144992 flock Species 0.000 description 1
- 210000004907 gland Anatomy 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000008844 regulatory mechanism Effects 0.000 description 1
- 230000000630 rising effect Effects 0.000 description 1
Classifications
-
- 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
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The application discloses a resource scheduling method based on an improved multi-neighborhood local enhancement search algorithm, which comprises the following steps: establishing a mapping relation between a resource scheduling scheme set facing a discrete workshop and a resource scheduling method solution space to generate a group of random populations; initializing a random population based on a double-rule coordination initialization method to obtain an initial population; according to the initial population, carrying out iterative updating of global search until a preset global search iteration termination condition is met, and outputting a final population optimal position; obtaining an optimal combined solution allocated to the corresponding work procedure according to the final population optimal position; the local search and the memory pool matrix are introduced in the global search, so that the global search capability and the local search depth of the algorithm are improved, the algorithm can be prevented from being converged prematurely, the method has stronger optimizing capability for production systems of different scales, and a high-quality resource scheduling scheme can be stably obtained.
Description
Technical Field
The application relates to a resource scheduling method based on an improved multi-neighborhood local enhancement search algorithm, and belongs to the technical field of production and manufacturing.
Background
The machine tool in China has large base number, more energy consumption and low energy efficiency, but has large energy-saving potential. The cost of researching and developing the energy-saving machine tool is high and the cost is low by only relying on process optimization, and compared with the workshop scheduling, the method can realize multi-index optimization in manufacturing production at lower cost by reasonably planning workshop resources. In addition, the diversified and personalized demands of users force the manufacturing industry to change to a multi-variety and small-scale manufacturing mode, the problem of scheduling in a standard discrete workshop eliminates the limitation of machine uniqueness in the processing process, meets the multi-variety and small-scale manufacturing demands of enterprises, and meets the demands of the contemporary discrete manufacturing mode better.
At present, most methods aiming at the resource allocation problem in the manufacturing process of enterprises are low in searching efficiency, single in energy-saving scheduling strategy, lack of a better initialization strategy, single in algorithm optimizing strategy and poor in self-adaptive capability, and the production efficiency of the enterprises is affected. The actual application capability of most of the existing resource coordination methods is not strong, the operation efficiency of enterprises cannot be improved, and the operation cost of the enterprises cannot be effectively reduced. Therefore, a high-efficiency and feasible scheduling method with strong robustness is researched to solve the problem of resource allocation in a production system, so that reasonable resource allocation is realized, the operation benefit of enterprises is improved, and the method has important theoretical value and practical significance.
The information disclosed in this background section is only for enhancement of understanding of the general background of the application and should not be taken as an acknowledgement or any form of suggestion that this information forms the prior art already known to a person of ordinary skill in the art.
Disclosure of Invention
The application aims to overcome the defects in the prior art, provides a resource scheduling method based on an improved multi-neighborhood local enhancement search algorithm, improves the global search capability and the local search depth of the algorithm by introducing local search and memory pool matrixes in global search, can avoid premature convergence of the algorithm, has stronger optimizing capability for production systems of different scales, and can stably obtain a high-quality resource scheduling scheme.
In order to achieve the above purpose, the application is realized by adopting the following technical scheme:
the application discloses a resource scheduling method based on an improved multi-neighborhood local enhancement search algorithm, which comprises the following steps:
establishing a mapping relation between a resource scheduling scheme set facing a discrete workshop and a resource scheduling method solution space, and generating a group of random populations, wherein one individual in the random populations corresponds to one resource scheduling scheme, and the resource scheduling scheme is a combination solution of workpiece procedure allocation;
initializing the random population based on a double-rule coordination initialization method to obtain an initial population;
according to the initial population, performing iterative updating of global search until a preset global search iteration termination condition is met, and outputting a final population optimal position;
obtaining an optimal combined solution allocated to the corresponding work procedure according to the final population optimal position;
wherein the iterative updating of the global search comprises the steps of:
according to the initial or last global updated population, carrying out iterative updating of local searching until a preset local searching iteration termination condition is met, and outputting the locally updated population and the optimal position of the population;
responding to the condition of meeting the preset global search iteration termination, and taking the locally updated population optimal position as the final population optimal position; otherwise, performing migration operation on the partially updated population based on a preset memory pool matrix and migration probability to obtain the migrated population; and carrying out position exchange on the population after the migration to obtain the population after the current global updating, and returning to the step of carrying out iterative updating of local searching for cyclic iteration.
Further, the dual rule coordination initialization method comprises the following steps:
each individual in the random population corresponds to a combination solution, an objective function of the combination solution is calculated, all the combination solutions are subjected to rapid non-dominated sorting and classification according to Pareto dominance relation, and a solution set of different Pareto grades is obtained;
calculating the extraction quantity of each combined solution of the solution sets of the first two Pareto grades;
and responding to the fact that the combined solutions in the solutions of the first two Pareto grades meet a preset extraction condition, extracting the combined solutions as individuals of the initial population until the extraction quantity is reached, ending the extraction, and finally generating the initial population.
Further, the expression of the extraction number of the combined solution is as follows:
where Dh represents the number of extractions of the combined solution of the h-th solution set, where h=1 represents a solution set of Pareto rank 1 and h=2 represents a solution set of Pareto rank 2; k (k) 2 Representing the total number of extractions of the combined solutions according to the crowdedness criteria; n is n 2 Representing the total number of combined solutions; n (N) h Representing the number of combined solutions in the h-th solution set;
the expression of the preset extraction condition is as follows:
wherein s is h,p Indicating the congestion degree of the p-th combined solution in the h-th solution set; s is S h Representing the h th solution setIs a congestion degree of (3); n (N) h Representing the number of combined solutions in the h-th solution set; z represents the Pareto scale.
Further, the iterative updating of the local search includes the steps of:
according to the initial or last global updated population, carrying out local enhancement information initialization, taking the maximum finishing time and the total processing energy consumption as objective functions, calculating the fitness value of each individual of the population, assigning the position of each individual as an initial individual optimal position, and assigning the position of the individual with the highest fitness value as an initial population optimal position;
based on a biological hormone feedback regulation mechanism and combining self-learning and social learning behaviors of examples, carrying out particle swarm local enhancement search and updating the speed and the position of an individual to obtain a current local updated population;
updating the optimal position of the individual and the optimal position of the population by calculating the fitness value of the individual after the current local update;
judging whether a preset local search iteration termination condition is met, outputting a population and a population optimal position after local updating in response to the preset local search iteration termination condition, and otherwise, returning to the step of particle swarm local enhancement searching for iterative updating.
Further, the updating of the individual optimal position and the population optimal position includes the following steps:
comparing the fitness value of the current locally updated individual with the fitness value of the initial or last locally updated individual optimal position, if the fitness value is better than the fitness value, assigning the current locally updated individual optimal position to the current locally updated individual optimal position, otherwise, reserving the individual optimal position;
and comparing each individual after the current local update with the fitness value of the optimal position of the population after the initial or last local update, if the fitness value is better than the fitness value, assigning the position of the individual after the current local update to the optimal position of the population after the current local update, otherwise, reserving the optimal position of the population.
Further, the fitness value is expressed as follows:
wherein C represents a fitness value; w (w) 1 Representing a first weight factor; w (w) 2 Representing a second weight factor; f (f) 1 An objective function representing an optimal maximum completion time; f (f) 2 An objective function representing the total energy consumption of the optimized process;
the objective function f 1 The expression of (2) is as follows:
f 1 ＝T max ＝max(T i ),i∈N
wherein T is max Representing the maximum finishing time of all the workpieces; t (T) i Indicating the finishing time of the workpiece i; n represents the total number of workpieces;
the objective function f 2 The expression of (2) is as follows:
wherein E represents the total energy consumption of the process; p (P) s (t) represents machine start-up power; y is Y i,j,m Representing a decision variable Y when the j-th process of the i-th workpiece is processed on machine m i,j,m 1, otherwise 0; p (P) u Representing machine standby power;a waiting time of machine m in a j-th process of machining an i-th workpiece; p (P) c Representing the machining power; n represents the total number of workpieces; m represents the total number of machines; o (O) i Indicating the total number of processes for the ith workpiece.
Further, in the particle swarm local enhancement search, the individual velocity and location update formula is as follows:
X l (t+1)＝X l (t)+V l (t+1)
wherein V is l (t+1) represents the speed of the first individual at the t+1st local iteration; v (V) l (t) represents the speed of the first individual at the t-th local iteration; g represents a threshold parameter; w (w) max Representing an inertial factor maximum; w (w) min Representing an inertia factor minimum; w (w) 0 Representing an inertial factor initial value; c 1 Representing a first self-learning behavioral parameter; r is (r) 1 Representing a second self-learning behavioral parameter; r is (r) 2 Representing a second social learning behavior parameter; p (P) l (t) represents an individual optimal location; p (P) g (t) represents a population optimal position; x is X l (t+1) represents the position of the first individual at the t+1st local iteration; x is X l (t) represents the position of the first individual at the t-th local iteration.
Further, the population after the migration comprises the following steps:
according to the population after local updating, performing memory matrix updating operation to obtain an updated memory matrix;
according to the updated memory Chi Juzhen, performing migration operation based on a preset migration probability to obtain a population after migration;
wherein, the memory pool matrix updating operation comprises the following steps:
responding to the iteration number of the global search being 1, and selecting s individuals before the fitness of the population after local updating to fill the memory pool matrix;
and in response to the iteration number of the global search being greater than 1, randomly selecting individuals with high adaptability of the locally updated population to replace individuals in the memory pool matrix.
Further, the expression of the migration operation is as follows:
in the method, in the process of the application,represents the kth+1st global iteration the first individual; LO represents the crossover operator; />Representing a kth global iteration a first individual; x (k) represents the randomly selected individuals in the updated memory matrix; rand represents a random number between (0, 1); pr represents the migration probability.
Further, the method for carrying out position exchange on the population after migration comprises the following two steps: OS sequence block exchange and OS element point exchange;
the steps of the OS sequence block exchange are as follows: according to the population after migration, two crossing positions are randomly selected, the two crossing positions are respectively located at the nearest end to form two element sequences, and the positions of the two element sequences are exchanged to obtain a first exchange population;
the steps of the OS element point exchange are as follows: and according to the first exchange population, two element points are randomly determined and exchanged to obtain the population after the current global update.
Compared with the prior art, the application has the beneficial effects that:
according to the resource scheduling method based on the improved multi-neighborhood local enhancement search algorithm, local search is introduced into global search, so that the global search capacity and the local search depth of the algorithm are improved; on the other hand, the memory pool matrix is introduced, so that the premature convergence of the algorithm can be avoided. The optimal combination solution obtained by the method can effectively shorten the production period, balance the machine load, reduce the processing energy consumption and improve the flexibility, the efficiency and the stability in the manufacturing process.
Drawings
FIG. 1 is a flow chart of a resource scheduling method based on an improved multi-neighborhood local enhanced search algorithm;
FIG. 2 is a schematic diagram of OS sequence block exchange provided by an embodiment;
FIG. 3 is a schematic diagram of OS element point exchange provided by an embodiment.
Detailed Description
The application is further described below with reference to the accompanying drawings. The following examples are only for more clearly illustrating the technical aspects of the present application, and are not intended to limit the scope of the present application.
The embodiment provides a resource scheduling method based on an improved multi-neighborhood local enhancement search algorithm, which comprises the following steps:
establishing a mapping relation between a resource scheduling scheme set facing a discrete workshop and a resource scheduling method solution space, and generating a group of random populations, wherein one individual in the random populations corresponds to one resource scheduling scheme, and the resource scheduling scheme is a combined solution of workpiece procedure allocation;
initializing a random population based on a double-rule coordination initialization method to obtain an initial population;
according to the initial population, carrying out iterative updating of global search until a preset global search iteration termination condition is met, and outputting a final population optimal position;
obtaining an optimal combined solution allocated to the corresponding work procedure according to the final population optimal position;
wherein the iterative updating of the global search comprises the steps of:
according to the population subjected to initial or last global updating, carrying out iterative updating of local searching until a preset local searching iteration termination condition is met, and outputting the population subjected to local updating and the optimal position of the population;
responding to the condition of meeting the preset global search iteration termination, and taking the locally updated population optimal position as the final population optimal position; otherwise, performing migration operation on the partially updated population based on a preset memory pool matrix and migration probability to obtain the migrated population; and (3) carrying out position exchange on the migrated population to obtain the population subjected to the current global updating, and returning to the step of carrying out iterative updating of local searching for cyclic iteration.
The technical conception of the application is as follows: on one hand, local search is introduced in global search, so that the global search capability and the local search depth of an algorithm are improved; on the other hand, the memory pool matrix is introduced, so that the premature convergence of the algorithm can be avoided. The optimal combination solution obtained by the method can effectively shorten the production period, balance the machine load, reduce the processing energy consumption and improve the flexibility, the efficiency and the stability in the manufacturing process.
As shown in fig. 1, the specific steps are as follows:
step 1: generating random populations
Establishing a mapping relation between a resource scheduling scheme set facing a discrete workshop and a resource scheduling method solution space to generate a group of random populations, wherein one individual in the random populations corresponds to one resource scheduling scheme, and the resource scheduling scheme is a combined solution for workpiece procedure allocation, and specifically comprises the following steps:
1.1 presetting algorithm parameters including population scale, maximum global search iteration number, maximum local search iteration number, memory matrix scale, migration probability and the like.
The solution quality of the scheduling optimization algorithm is affected by the number of individuals and the number of iterations of the algorithm. Selecting an excessive number of individuals and number of iterations increases the probability of searching for the optimal solution, but increases computation time significantly. Selecting too small an individual number and number of iterations reduces the computational effort, but the quality of the solution decreases. Considering the superiority of calculation time and solution comprehensively, the population scale range in the embodiment is between 30 and 50, the maximum global search iteration frequency range is selected between 50 and 600 times, the maximum local search iteration frequency range is selected between 30 and 50 times, and specific numerical values can be determined according to actual requirements.
The primary problem of establishing the resource scheduling optimization problem facing the discrete workshop manufacturing process is to establish a mapping relation between a resource scheduling scheme set facing the discrete workshop manufacturing process and a scheduling optimization method solution space, wherein the scheduling scheme is a combined solution of workpiece processes, and the solution space is a set of all feasible scheduling schemes. Coding is the conversion of a viable scheduling scheme into a sequence of process sequences, i.e., chromosomes, in which each number represents a workpiece and the frequency of occurrence of each number represents the process that the workpiece contains. Assuming that an order contains 3 work pieces, wherein the 1 st work piece contains 3 work pieces, the 2 nd work piece contains 2 work pieces, and the 3 rd work piece contains 4 work pieces, and 9 work pieces are total, the chromosome contains 9 genes (each work piece represents one gene), and is composed of 31, 2 and 4 3, and this process is a coded process. Decoding is the conversion of a chromosome into a scheduling scheme, i.e., the position of the same number on the chromosome represents the processing sequence of the process represented by that number on the scheduling plan. Continuing with the above assumption, in the chromosome, 1 st 1 denotes the first process of the workpiece 1, 1 st 2 denotes the first process of the workpiece 2, and 2 nd 2 denotes the 2 nd process of the workpiece 2, which is the decoding process.
A set of random populations is generated, each individual in the population comprising both a location and a velocity attribute.
Wherein for each row of the position matrix one feasible solution is represented. The number of rows is the number of individuals in the population, and the number of columns is the sum of the number of work-piece steps. Sequencing random numbers of each row of arrays in the position matrix from small to large by using a sort function to obtain a sequencing array and a subscript array A;
each element of the index group a is converted into a procedure, and the conversion procedure is shown as follows:
wherein A is]An a-th element representing the index group a; o (O) 1 Representing the total number of working procedures of the 1 st workpiece; o (O) i,j A j-th step of expressing an i-th workpiece; o (O) i Indicating the total number of processes for the ith workpiece.
For each solution, machine allocation is carried out according to a machine allocation mechanism, a scheduling result corresponding to the solution is obtained, and two targets in the text can be known according to the scheduling result: maximum finishing time and total processing energy consumption.
Based on the maximum finishing time and the total processing energy consumption, the individual fitness value is expressed as follows:
wherein, C is a tableShowing the fitness value; w (w) 1 Representing a first weight factor; w (w) 2 Representing a second weight factor; f (f) 1 An objective function representing an optimal maximum completion time; f (f) 2 An objective function representing the total energy consumption of the optimized process;
objective function f 1 The expression of (2) is as follows:
f 1 ＝T max ＝max(T i ),i∈N
wherein T is max Representing the maximum finishing time of all the workpieces; t (T) i Indicating the finishing time of the workpiece i; n represents the total number of workpieces;
objective function f 2 The expression of (2) is as follows:
wherein E represents the total energy consumption of the process; p (P) s (t) represents machine start-up power; y is Y i,j,m Representing a decision variable Y when the j-th process of the i-th workpiece is processed on machine m i,j,m 1, otherwise 0; p (P) u Representing machine standby power;a waiting time of machine m in a j-th process of machining an i-th workpiece; p (P) c Representing the machining power; n represents the total number of workpieces; m represents the total number of machines; o (O) i Representing the total number of processes of the ith workpiece;
step 2: population initialization
Initializing a random population based on a double-rule coordination initialization method to obtain an initial population;
the dual rule coordination initialization method comprises the following steps:
each individual in the random population corresponds to a combination solution, an objective function of the combination solution is calculated, and all the combination solutions are subjected to rapid non-dominated sorting and classification according to Pareto dominance relation, so that solution sets of different Pareto grades are obtained;
calculating the extraction quantity of each combined solution of the solution sets of the first two Pareto grades; the expression of the number of decimations of the combined solution is as follows:
wherein D is h Representing the number of decimations of the combined solution of the h-th solution set, where h=1 represents a solution set of Pareto rank 1 and h=2 represents a solution set of Pareto rank 2; k (k) 2 Representing the total number of extractions of the combined solutions according to the crowdedness criteria; n is n 2 Representing the total number of combined solutions; n (N) h Representing the number of combined solutions in the h-th solution set.
Determining the degree of congestion s of the combined solutions in the solutions of the first two Pareto classes i,j And the sum of the crowding degree S i ；
And responding to the fact that the combined solutions in the first two Pareto-level solution sets meet a preset extraction condition, extracting the combined solutions as individuals of the initial population until the extraction quantity of the combined solutions reaching the Pareto front solution set or the rest solution sets is ended.
The expression of the preset extraction condition is as follows:
wherein s is h,p Indicating the congestion degree of the p-th combined solution in the h-th solution set; s is S h Representing the crowding degree of the h solution set; n (N) h Representing the number of combined solutions in the h-th solution set; z represents Pareto grade;
the first two Pareto-level solution sets are combined to extract a combined solution, and finally an initial population is generated.
Step 3: global search
According to the initial population, carrying out iterative updating of global search until a preset global search iteration termination condition is met, and outputting the final optimal position of the population, wherein the method comprises the following specific steps:
3.1 local search
According to the population subjected to initial or last global updating, carrying out iterative updating of local searching until a preset local searching iteration termination condition is met, and outputting the population subjected to local updating and the optimal position of the population;
3.1.1 local enhancement information initialization.
According to the initial or last global updated population, carrying out local enhancement information initialization, taking the maximum finishing time and the total processing energy consumption as objective functions, calculating the fitness value of each individual of the population, assigning the position of each individual as an initial individual optimal position, and assigning the position of the individual with the highest fitness value as an initial population optimal position;
3.1.2 particle swarm local enhanced search
Based on a biological hormone feedback regulation mechanism and combining self-learning and social learning behaviors of examples, carrying out particle swarm local enhancement search and updating the speed and the position of an individual to obtain a current local updated population.
The locally enhanced search operation of the bio-hormone feedback regulation mechanism is designed as follows:
the biological hormone feedback regulation mechanism has the characteristics of high efficiency, specificity and the like, and can rapidly maintain the internal environment to keep stable. The study of the principle of hormone feedback regulation by Farhy proves that the hormone regulation mechanism accords with Hill function, and the function comprises two parts, namely an ascending function and a descending function, and the expression is as follows.
Wherein F is up (q) represents a rising function of hormone q; f (F) down (q) represents a decreasing function of hormone q; q represents the hormone concentration; l represents the concentration threshold of the hormone, and L>0; d represents HILL coefficient, and d>0；
From the above equation, it can be analyzed that the derivative value of the curve at each point is determined by both d and L.
Meanwhile, farhy made a related experiment on the hormone feedback regulator according to the above, and based on the final experimental result, farhy indicated that if hormone 2 regulates the gland separating hormone 1, the relation between the secretion rate of hormone 1 and the concentration of hormone 2 is shown as formula (19).
In the method, in the process of the application,indicating an ascending secretion rate of hormone 1; />Representing a decreased secretion rate of hormone 1; />Represents the initial secretion rate of hormone 1; q 0 Indicating the initial concentration, the initial concentrations of hormone 1 and hormone 2 are consistent; q 2 Represents the concentration of hormone 2; d represents a HILL coefficient;
in the PSO operator, the inertial factor is often designed by adopting a line function or a constant function, and due to the linear characteristic, the search is difficult to avoid being trapped in a local optimal neighborhood structure. According to the foregoing description, the bio-hormone feedback adjustment function has monotonicity and nonlinearity, and can better adjust the step size of the local search, based on which the inertia factor w is designed as follows:
wherein t represents the number of local iterations; g represents a threshold parameter; w (w) max Representing an inertial factor maximum; w (w) min Representing an inertia factor minimum; w (w) 0 Representing an inertial factor initial value; d represents HILL coefficient, and d>0。
HO et al point out the self-learning behavior of the flock (c 1 *r 1 ) And social learning behavior (c) 2 *(1-r 2 ) Is random and independent of each other c 1 +c 2 ＝1，c 1 *r 1 Affecting the step length of the individual flying to the history of the individual, c 2 *r 2 The optimal step length of the individual flying to the group history is influenced, and the adjustment of two learning behaviors has a certain influence on the quality of the solution.
Therefore, based on a bio-hormone feedback regulation mechanism, combining two learning behaviors of particles, the LESO of this section is proposed, and in the particle swarm local enhancement search, the speed and the position of an individual update formula is as follows:
X l (t+1)＝X l (t)+V l (t+1)
wherein V is l (t+1) represents the speed of the first individual at the t+1st local iteration; v (V) l (t) represents the speed of the first individual at the t-th local iteration; g represents a threshold parameter; w (w) max Representing an inertial factor maximum; w (w) min Representing an inertia factor minimum; w (w) 0 Representing an inertial factor initial value; c 1 Representing a first self-learning behavioral parameter; r is (r) 1 Representing a second self-learning behavioral parameter; r is (r) 2 Representing a second social learning behavior parameter; p (P) l (t) represents an individual optimal location; p (P) g (t) represents a population optimal position; x is X l (t+1) represents the position of the first individual at the t+1st local iteration; x is X l (t) represents the position of the first individual at the t-th local iteration.
3.1.3 updating the optimal position of the individual and the optimal position of the population by calculating the fitness value of the individual after the current local update,
updating the individual optimal position and the population optimal position comprises the following steps:
comparing the fitness value of the current locally updated individual with the fitness value of the initial or last locally updated individual optimal position, if the fitness value is better than the fitness value, assigning the current locally updated individual optimal position to the current locally updated individual optimal position, otherwise, reserving the individual optimal position;
and comparing each individual after the current local update with the fitness value of the optimal position of the population after the initial or last local update, if the fitness value is better than the fitness value, assigning the position of the individual after the current local update to the optimal position of the population after the current local update, otherwise, reserving the optimal position of the population.
And 3.1.4, judging whether a preset local search iteration termination condition is met.
Judging whether a preset local search iteration termination condition is met, outputting a population and a population optimal position after local updating in response to the preset local search iteration termination condition, and otherwise, returning to the step of local enhancement search of the 31.2 particle swarm to carry out iterative updating.
The preset local search iteration termination condition in this embodiment is the maximum local search iteration number preset in step 1.1.
3.2 judging whether the preset global search iteration termination condition is met
And responding to the preset global search iteration termination condition, and taking the locally updated population optimal position as a final population optimal position.
In this embodiment, the preset global search iteration termination condition is the maximum global search iteration number preset in step 1.1.
3.3 population migration
Performing migration operation on the population after local updating based on a preset memory pool matrix and migration probability to obtain the population after migration, and specifically comprising the following steps of:
according to the population after local updating, performing memory matrix updating operation to obtain an updated memory matrix;
according to the updated memory Chi Juzhen, performing migration operation based on the preset migration probability to obtain a population after migration;
the memory matrix updating operation includes the following steps:
responding to the iteration number of the global search being 1, and selecting s individuals before the fitness of the population after local updating to fill the memory pool matrix;
and in response to the iteration number of the global search being greater than 1, randomly selecting individuals with high adaptability of the locally updated population to randomly replace individuals in the memory pool matrix.
The representation of the memory cell matrix is shown below.
In the formula, the first column of the matrix is a number, s represents the number of solutions in the matrix, x (k) represents the spatial position of the k-th iteration solution of the global search, and u represents the number of features. The memory matrix pool is used for dynamically storing the high-quality solution obtained by the algorithm, and the expression of the migration operation is as follows:
in the method, in the process of the application,represents the kth+1st global iteration the first individual; LO represents the crossover operator; />Representing a kth global iteration a first individual; x (k) represents the randomly selected individuals in the updated memory matrix; rand represents a random number between (0, 1); pr represents the migration probability.
3.4 location switching
And (3) performing position exchange on the migrated population to obtain the population subjected to the current global updating, and returning to the step (3.1) of performing iterative updating of local searching to perform loop iteration.
The method for carrying out position exchange on the migrated population comprises the following two steps: OS sequence block exchange and OS element point exchange;
the steps of OS sequence block exchange are: according to the migrated population, randomly selecting two crossing positions, wherein the two crossing positions are respectively positioned at the nearest end to form two element sequences, and exchanging the positions of the two element sequences to obtain a first exchange population; as particularly shown in fig. 2.
The steps of OS element point exchange are: according to the first exchange population, two element points are randomly determined and exchanged to obtain a population after current global updating; as particularly shown in fig. 3.
Step 4: obtaining an optimal combined solution allocated to the corresponding work procedure according to the final population optimal position;
in summary, the application firstly analyzes the characteristics of the energy-saving scheduling problem of the multi-target discrete workshops, and establishes the mapping relation between the resource scheduling scheme set facing the discrete workshops and the solution space of the resource scheduling method; secondly, a machine greedy allocation mechanism is provided, three allocation criteria are designed, and the search dimension of the multi-target discrete workshop energy-saving scheduling problem is reduced; finally, a multi-neighborhood local enhancement search algorithm is provided, local search is introduced in global search, and global search capacity and local search depth of the algorithm are improved; and meanwhile, a memory pool matrix is introduced, so that the premature convergence of an algorithm can be avoided. The optimal combination solution obtained by the method can effectively shorten the production period, balance the machine load, reduce the processing energy consumption and improve the flexibility, the efficiency and the stability in the manufacturing process.
It will be appreciated by those skilled in the art that embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
The foregoing is only a preferred embodiment of the application, it being noted that: it will be apparent to those skilled in the art that various modifications and adaptations can be made without departing from the principles of the present application, and such modifications and adaptations are intended to be comprehended within the scope of the application.
Claims (10)
1. The resource scheduling method based on the improved multi-neighborhood local enhancement search algorithm is characterized by comprising the following steps:
establishing a mapping relation between a resource scheduling scheme set facing a discrete workshop and a resource scheduling method solution space, and generating a group of random populations, wherein one individual in the random populations corresponds to one resource scheduling scheme, and the resource scheduling scheme is a combination solution of workpiece procedure allocation;
initializing the random population based on a double-rule coordination initialization method to obtain an initial population;
according to the initial population, performing iterative updating of global search until a preset global search iteration termination condition is met, and outputting a final population optimal position;
obtaining an optimal combined solution allocated to the corresponding work procedure according to the final population optimal position;
wherein the iterative updating of the global search comprises the steps of:
according to the initial or last global updated population, carrying out iterative updating of local searching until a preset local searching iteration termination condition is met, and outputting the locally updated population and the optimal position of the population;
responding to the condition of meeting the preset global search iteration termination, and taking the locally updated population optimal position as the final population optimal position; otherwise, performing migration operation on the partially updated population based on a preset memory pool matrix and migration probability to obtain the migrated population; and carrying out position exchange on the population after the migration to obtain the population after the current global updating, and returning to the step of carrying out iterative updating of local searching for cyclic iteration.
2. The resource scheduling method based on the improved multi-neighborhood local enhanced search algorithm according to claim 1, wherein the steps of the dual rule coordination initialization method are as follows:
each individual in the random population corresponds to a combination solution, an objective function of the combination solution is calculated, all the combination solutions are subjected to rapid non-dominated sorting and classification according to Pareto dominance relation, and a solution set of different Pareto grades is obtained;
calculating the extraction quantity of each combined solution of the solution sets of the first two Pareto grades;
and responding to the fact that the combined solutions in the solutions of the first two Pareto grades meet a preset extraction condition, extracting the combined solutions as individuals of the initial population until the extraction quantity is reached, ending the extraction, and finally generating the initial population.
3. The resource scheduling method based on the improved multi-neighborhood local enhancement search algorithm according to claim 2, wherein the expression of the extraction number of the combined solution is as follows:
where Dh represents the number of extractions of the combined solution of the h-th solution set, where h=1 represents a solution set of Pareto rank 1 and h=2 represents a solution set of Pareto rank 2; k (k) 2 Representing the total number of extractions of the combined solutions according to the crowdedness criteria; n is n 2 Representing the total number of combined solutions; n (N) h Representing the number of combined solutions in the h-th solution set;
the expression of the preset extraction condition is as follows:
wherein s is h,p Indicating the congestion degree of the p-th combined solution in the h-th solution set; s is S h Representing the crowding degree of the h solution set; n (N) h Representing the number of combined solutions in the h-th solution set; z represents the Pareto scale.
4. The resource scheduling method based on the improved multi-neighborhood local enhanced search algorithm according to claim 1, wherein the iterative updating of the local search comprises the steps of:
according to the initial or last global updated population, carrying out local enhancement information initialization, taking the maximum finishing time and the total processing energy consumption as objective functions, calculating the fitness value of each individual of the population, assigning the position of each individual as an initial individual optimal position, and assigning the position of the individual with the highest fitness value as an initial population optimal position;
based on a biological hormone feedback regulation mechanism and combining self-learning and social learning behaviors of examples, carrying out particle swarm local enhancement search and updating the speed and the position of an individual to obtain a current local updated population;
updating the optimal position of the individual and the optimal position of the population by calculating the fitness value of the individual after the current local update;
judging whether a preset local search iteration termination condition is met, outputting a population and a population optimal position after local updating in response to the preset local search iteration termination condition, and otherwise, returning to the step of particle swarm local enhancement searching for iterative updating.
5. The resource scheduling method based on the improved multi-neighborhood local enhancement search algorithm according to claim 4, wherein updating the individual optimum position and the population optimum position comprises the steps of:
comparing the fitness value of the current locally updated individual with the fitness value of the initial or last locally updated individual optimal position, if the fitness value is better than the fitness value, assigning the current locally updated individual optimal position to the current locally updated individual optimal position, otherwise, reserving the individual optimal position;
and comparing each individual after the current local update with the fitness value of the optimal position of the population after the initial or last local update, if the fitness value is better than the fitness value, assigning the position of the individual after the current local update to the optimal position of the population after the current local update, otherwise, reserving the optimal position of the population.
6. The resource scheduling method based on the improved multi-neighborhood local enhanced search algorithm according to claim 4, wherein the expression of the fitness value is as follows:
wherein C represents a fitness value; w (w) 1 Representing a first weight factor; w (w) 2 Representing a second weight factor; f (f) 1 An objective function representing an optimal maximum completion time; f (f) 2 Target for optimizing total energy consumption of processingA function;
the objective function f 1 The expression of (2) is as follows:
f 1 ＝T max ＝max(T i ),i∈N
wherein T is max Representing the maximum finishing time of all the workpieces; t (T) i Indicating the finishing time of the workpiece i; n represents the total number of workpieces;
the objective function f 2 The expression of (2) is as follows:
wherein E represents the total energy consumption of the process; p (P) s (t) represents machine start-up power; y is Y i,j,m Representing a decision variable Y when the j-th process of the i-th workpiece is processed on machine m i,j,m 1, otherwise 0; p (P) u Representing machine standby power;a waiting time of machine m in a j-th process of machining an i-th workpiece; p (P) c Representing the machining power; n represents the total number of workpieces; m represents the total number of machines; o (O) i Indicating the total number of processes for the ith workpiece.
7. The resource scheduling method based on the improved multi-neighborhood local enhancement search algorithm according to claim 4, wherein in the particle swarm local enhancement search, the speed and the position of the individual update formula are as follows:
X l (t+1)＝X l (t)+V l (t+1)
wherein V is l (t+1) represents the speed of the first individual at the t+1st local iteration; v (V) l (t) represents the speed of the first individual at the t-th local iteration;g represents a threshold parameter; w (w) max Representing an inertial factor maximum; w (w) min Representing an inertia factor minimum; w (w) 0 Representing an inertial factor initial value; c 1 Representing a first self-learning behavioral parameter; r is (r) 1 Representing a second self-learning behavioral parameter; r is (r) 2 Representing a second social learning behavior parameter; p (P) l (t) represents an individual optimal location; p (P) g (t) represents a population optimal position; x is X l (t+1) represents the position of the first individual at the t+1st local iteration; x is X l (t) represents the position of the first individual at the t-th local iteration.
8. The resource scheduling method based on the improved multi-neighborhood local enhancement search algorithm according to claim 1, wherein the obtained migrated population comprises the steps of:
according to the population after local updating, performing memory matrix updating operation to obtain an updated memory matrix;
according to the updated memory Chi Juzhen, performing migration operation based on a preset migration probability to obtain a population after migration;
wherein, the memory pool matrix updating operation comprises the following steps:
responding to the iteration number of the global search being 1, and selecting s individuals before the fitness of the population after local updating to fill the memory pool matrix;
and in response to the iteration number of the global search being greater than 1, randomly selecting individuals with high adaptability of the locally updated population to replace individuals in the memory pool matrix.
9. The resource scheduling method based on the improved multi-neighborhood local enhancement search algorithm according to claim 8, wherein the expression of the migration operation is as follows:
in the method, in the process of the application,represents the kth+1st global iteration the first individual; LO represents the crossover operator; />Representing a kth global iteration a first individual; x (k) represents the randomly selected individuals in the updated memory matrix; rand represents a random number between (0, 1); pr represents the migration probability.
10. The resource scheduling method based on the improved multi-neighborhood local enhancement search algorithm according to claim 1, wherein the performing of a position exchange on the migrated population comprises two steps: OS sequence block exchange and OS element point exchange;
the steps of the OS sequence block exchange are as follows: according to the population after migration, two crossing positions are randomly selected, the two crossing positions are respectively located at the nearest end to form two element sequences, and the positions of the two element sequences are exchanged to obtain a first exchange population;
the steps of the OS element point exchange are as follows: and according to the first exchange population, two element points are randomly determined and exchanged to obtain the population after the current global update.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311274018.6A CN117217487A (en) | 2023-09-28 | 2023-09-28 | Resource scheduling method based on improved multi-neighborhood local enhancement search algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311274018.6A CN117217487A (en) | 2023-09-28 | 2023-09-28 | Resource scheduling method based on improved multi-neighborhood local enhancement search algorithm |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117217487A true CN117217487A (en) | 2023-12-12 |
Family
ID=89036983
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311274018.6A Pending CN117217487A (en) | 2023-09-28 | 2023-09-28 | Resource scheduling method based on improved multi-neighborhood local enhancement search algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117217487A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118521139A (en) * | 2024-07-24 | 2024-08-20 | 苏州简诺科技有限公司 | System resource demand planning method and system based on artificial intelligence |
-
2023
- 2023-09-28 CN CN202311274018.6A patent/CN117217487A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118521139A (en) * | 2024-07-24 | 2024-08-20 | 苏州简诺科技有限公司 | System resource demand planning method and system based on artificial intelligence |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110796355B (en) | Flexible job shop scheduling method based on dynamic decoding mechanism | |
| CN110598941A (en) | Bionic strategy-based dual-target scheduling method for particle swarm optimization manufacturing system | |
| CN117217487A (en) | Resource scheduling method based on improved multi-neighborhood local enhancement search algorithm | |
| CN101901425A (en) | Flexible job shop scheduling method based on multi-species coevolution | |
| CN108460463B (en) | High-end equipment assembly line production scheduling method based on improved genetic algorithm | |
| CN111401693B (en) | Flexible workshop scheduling optimization method and system with robot transportation | |
| CN105373845A (en) | Hybrid intelligent scheduling optimization method of manufacturing enterprise workshop | |
| CN110738365A (en) | flexible job shop production scheduling method based on particle swarm optimization | |
| CN113569483A (en) | Method for solving multi-target flexible job shop scheduling based on artificial bee colony algorithm | |
| CN112668789A (en) | Self-adaptive batch scheduling method for flexible operation workshop preparation process | |
| CN113901728B (en) | Computer second-class assembly line balance optimization method based on migration genetic algorithm | |
| CN113821972A (en) | Multi-robot collaborative assembly line balancing method considering energy efficiency | |
| CN115730799A (en) | Method, system and equipment for scheduling production tasks of flexible assembly job workshop | |
| CN115907399A (en) | Intelligent scheduling method for discrete manufacturing flexible production of electronic product | |
| CN110490446A (en) | A kind of modular process recombination method based on improved adaptive GA-IAGA | |
| CN114648232A (en) | Cloud resource flexible job scheduling method based on improved chimpanzee optimization algorithm | |
| CN117555305B (en) | NSGAII-based multi-target variable sub-batch flexible workshop job scheduling method | |
| CN111210125A (en) | Multi-target workpiece batch scheduling method and device based on historical information guidance | |
| CN112631214B (en) | Flexible job shop batch scheduling method based on improved invasive weed optimization algorithm | |
| CN116985146B (en) | Robot parallel disassembly planning method for retired electronic products | |
| CN116501272A (en) | 3D printing order task scheduling method based on improved genetic algorithm | |
| CN114089699B (en) | Method for solving mixed flow shop scheduling based on cell type membrane calculation model | |
| CN114676987B (en) | Intelligent flexible job shop active scheduling method based on hyper-heuristic algorithm | |
| CN115456268A (en) | Guide roller manufacturing resource optimal allocation method, device, equipment and medium | |
| CN115700647A (en) | Workshop flexible operation scheduling method based on tabu search genetic algorithm |
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 |