CN112396214B - Method and system for smelting workshop scheduling based on improved teaching optimization algorithm - Google Patents

Abstract

The application provides a smelting workshop scheduling method and system based on an improved teaching optimization algorithm, and relates to the technical field of workshop scheduling optimization. The application provides an energy efficiency balance optimization model considering the dynamic arrival time of raw materials, an original teaching optimization algorithm is innovated for solving the model, class initialization, teaching links, self-adaptive learning links, bernoulli self-learning links and other aspects are respectively improved and innovated, the local innovation sub-process is incorporated into the framework of the improved teaching optimization algorithm by combining the structural characteristics of the model, the improved algorithm can more objectively describe the teaching links, the learning step length of class students is adaptively adjusted at the later stage of the algorithm, the self-learning ability of individual class students is improved, the approximate optimal solution of the energy efficiency balance optimization model can be solved within reasonable solving time, and certain theoretical guidance and support are provided for the production practice of enterprises.

Description

Method and system for smelting workshop scheduling based on improved teaching optimization algorithm

technical field

The invention relates to the technical field of workshop scheduling optimization, in particular to a smelting workshop scheduling method and system based on an improved teaching optimization algorithm.

Background technique

In the high-end equipment precision casting production line (such as precision aviation parts, heavy equipment structural parts, etc.), the raw materials of general metals such as aluminum, iron and some rare earth elements may not be fully prepared at the beginning of the production cycle. In most cases, raw materials may arrive at the melt shop at different times. Through information technology such as the Internet of Things, the dynamic arrival time of raw materials will be known by the workshop management system at the beginning of the production cycle. At the same time, the sales system will also provide the delivery time of the smelting workpiece order to the workshop management system. The smelting workshop operation is a key link in the energy consumption of the entire high-end equipment precision casting process, which has typical characteristics such as long energy consumption cycle, complex energy consumption composition, and huge total energy consumption. Reducing the energy consumption level of the melt shop operation can effectively reduce the energy cost of the investment casting production line. The operation of the smelting workshop is carried out around various smelting furnaces. The working process of the smelting furnace is generally divided into steps such as preheating, vibration feeding, smelting, refining, casting, and cooling. After analyzing the specific melting temperature curve, it can be found that after the casting is completed, the furnace still maintains a certain temperature and drops to the initial temperature after a period of cooling. How to solve the energy-saving production scheduling problem in the smelting workshop operation is an urgent problem to be solved.

The existing energy-saving production scheduling problem mainly uses the machine shutdown and restart strategy to reduce the energy consumption during the idle period of the production process, without considering the waste heat utilization phenomenon that may exist in the precision investment casting process of high-end equipment, it is difficult to use the existing Internet of Things technology Carry out effective energy efficiency balance optimization management.

The patent of the present invention aims to provide an improved teaching optimization algorithm for solving the energy efficiency balance optimization problem of precision furnace casting with the dynamic arrival time of raw materials and the machine sleep strategy. The steps of the existing teaching optimization algorithm are as follows:

(1) Initialize the number of students and the number of subjects;

(2) Randomly initialize the achievement information of each individual subject of the student and calculate the comprehensive achievement of each individual student;

(3) Take the student with the best comprehensive grade as the teacher and set the individual student with the average grade as an equal student individual;

(4) Use the information of equal individual students and individual teachers to iterate the teaching process for individual students;

(5) Use the individual information of any other student in the classroom to iterate the learning process for each individual student;

(6) Update the grades and comprehensive grades of teachers and equal students in each subject;

(7) Judging whether the iteration condition is satisfied, if so, go to step (4), otherwise stop the iteration, and output the current optimal solution.

The existing teaching optimization has shortcomings in randomizing the initial population, teaching links, learning links, etc., resulting in low search efficiency, and cannot balance the relationship between the diversity and convergence of the algorithm population.

Contents of the invention

(1) Solved technical problems

Aiming at the deficiencies of the existing technology, the present invention provides a smelting workshop scheduling method and system based on an improved teaching optimization algorithm, which solves the problem that the existing scheduling algorithm cannot utilize the waste heat that may exist in the precision investment casting process of high-end equipment.

(2) Technical solution

To achieve the above object, the present invention is achieved through the following technical solutions:

In the first aspect, a smelting workshop scheduling method based on an improved teaching optimization algorithm is provided, the method includes:

S1. Obtain scheduling information; set the total number of iterations, and initialize the number of iterations it=0;

S2. Randomly initialize the grades of 80% of individual students in each subject in the class, and randomly initialize the grades of each subject of the remaining 20% of individual students based on the principle of fully utilizing the waste heat of the furnace, and construct the initial grades of each subject of individual students in the class. The quantity is the same as the quantity of parts;

S3. Based on the individual students' initial grades in each subject, calculate the comprehensive grades of each individual student in the class, and construct the initial class grades;

S4. Take the individual student with the best overall score in the current class as the individual teacher;

The individual student whose comprehensive score is the average of each subject in the current class is regarded as an equal individual student;

Take the student individual whose comprehensive score is the median in the current class as the median individual student;

S5. Based on the current individual teacher, average student individual and median student individual, carry out the teaching process for the current class, and update the current class grade;

S6. Use the grades of any individual student in the class to perform adaptive learning for the individual students in the class, and update the current grades of the class again;

S7. Execute the self-learning link of individual students in the current class, and update the current class grades to form the class grades after the it-th iteration;

S8. Judging whether the number of iterations it reaches the total number of iterations Mit, if so, output the current optimal student individual score, and convert it into a scheduling sequence through the decoding algorithm; otherwise, update the current individual teacher, equal student individual and median student individual Subject grades, iteration times, and return to S5.

Further, the scheduling information includes:

Parts set Ω＝{J ₁ ,…,J _j ,…,J _N };

A collection of component dynamic arrival times

The warm-up time t _p required by the parts;

Melting time π={s ₁ ,…,s _j ,…,s _N };

Refining time ω={p ₁ ,...,p _j ,...p _N }.

Further, in the S2, randomly initialize the grades of 80% of the individual students in the class, and randomly initialize the grades of the remaining 20% of the individual students based on the principle of fully utilizing the waste heat of the furnace, and construct the initial grades of the individual students in the class. The specific steps for the grades are:

S201, randomly generating a random number between [0,1] as the student's individual random score;

S202, 20% of the students' individual random scores are decoded by the decoding algorithm to obtain the corresponding scheduling sequence, which is recorded as N is the number of parts;

S203. Set j=1, and start the smelting operation of part J _j at the time of max{t _p , r _j }, record the completion time of the refining operation of part J _j as C _j ;

t _p represents the warm-up time required by the component, and r _j represents the dynamic arrival time of the jth component;

S204. Determine whether the arrival time of the remaining parts does not exceed C _j , if yes, arrange the part with the largest sum of smelting time and refining time among all parts whose arrival time does not exceed C _j to start the smelting operation, and the refining operation is completed within the time Denote as C _j+1 ;

S205, judge whether j is less than N-1, if it is true, set j=j+1 and return to S204; otherwise, use the decoding algorithm to reverse the current scheduling sequence again, and obtain the students corresponding to 20% of the students' individual random scores The individual's initial grades of each subject, together with 80% of the individual random grades of the students, constitute the initial grades of each subject of the individual students in the class.

Further, S5, performing the teaching process on the current class based on the current individual teacher, average student individual and median student individual, and updating the current class grade, specifically includes the following steps:

S501. Initialize i=1, it=0; that is, when performing the first iteration, iterate based on the initial class grade;

The grades and comprehensive grades of the individual students in the it-th generation class are:

Among them, f(X ⁱ [it]) represents the initial comprehensive score of the i-th individual student in the class of the it-th generation; i=1,...,NP;

X ^NP [it] represents the NP-th student individual in the class of the it-th generation;

Indicates the initial score of the d-th subject of the NP-th individual student in the class of the it-th generation;

The grades and comprehensive grades of individual teachers in each subject are as follows:

The median grades of each subject and the comprehensive score of individual students are:

The scores of each subject and the comprehensive score of the average student are:

S502, let j=1;

S503. Let rand _i =rand(0,1), TF _i =1+rand(0,1),

And construct the grades of the first class of new student individual subject j:

S504. Let rand _i =rand(0,1) and TF _i =1+rand(0,1),

And construct the grades of the jth subject of the second type of new student individual; the construction formula is:

S505. Determine whether j≤N is established, if established, then set j=j+1 and turn to S503; otherwise, calculate the comprehensive scores of the two new types of student individuals New1X ⁱ [it] and New2X ⁱ [it] respectively, and Keep the one with the best comprehensive score among X ⁱ [it], New1X ⁱ [it] and New2X ⁱ [it] as X ⁱ [it+1];

S506. Determine whether i≤NP holds true, if true, set i=i+1, update individual teachers, average students, and median students, and turn to S502; otherwise, the teaching link is terminated.

Further, said S6, using the scores of any individual student in the class to perform adaptive learning for the individual student in the class, and updating the current class score again, specifically includes the following steps:

S601, setting i=1;

S602. Select any individual student who is different from Xi [it] as X ^k [it], and ^set j=1;

S603. Order Where θ>1, and construct the grade of the jth subject of the new individual student:

S604. Determine whether j≤N is true, if true, set j=j+1, and turn to S603; otherwise, calculate the comprehensive score of the new student individual NewX ⁱ [it], and compare ^Xi [it] and NewX The one with the best overall score in ⁱ [it] is reserved as X ⁱ [it+1];

S605. Determine whether i≤NP holds true, if true, set i=i+1, and go to S602; otherwise, the adaptive learning link is terminated.

Further, said S7, executing the self-learning link of individual students in the current class, and updating the current class grades to form the class grades after the it iteration, the specific steps are:

S701: set i=1,

S702, setting j=1;

S703. Let BR _i =rand(0,1), and construct the grade of subject j of a new individual student:

S704. Determine whether j≤N is true, if true, set j=j+1 and go to S703; otherwise, calculate the comprehensive score of the new student individual NewBX ⁱ [it], and compare ^Xi [it] and NewBX ⁱ The one with the best comprehensive score in [it] is reserved as X ⁱ [it+1];

S705: Determine whether i≤NP holds true, if true, set i=i+1, and go to S702; otherwise, the Bernoulli self-learning learning link is terminated.

Further, the calculation formula of the comprehensive score of the individual student is:

Among them, C _max represents the manufacturing span of the casting system, ∑E represents the total energy consumption of the casting system, the lower bounds of the manufacturing span and total energy consumption of the LB _C and LB _E casting systems, w _C and w _E are the manufacturing span and total energy consumption, respectively the weight of energy;

Among them, s _j represents the melting time of the jth component;

p _j represents the refining time of the jth component;

α is the utilization rate of waste heat, which means 0<α<1;

Δt _{j, j+1} represents the interval time between two parts processing;

t _p represents the warm-up time required by the parts

χ represents the cooling timeout threshold;

P _s represents the smelting power;

P _p represents the refining power;

P _pre represents the preheating power.

Further, the decoding algorithm is used to convert the grades of the it-th generation of individual students into a scheduling sequence, and the specific steps are:

Step 1: Make a one-to-one correspondence between the grades of individual students in each subject and the serial numbers of parts;

Step 2: Sort the serial numbers of the parts in ascending order according to the grades of individual students in each subject to obtain a new sorting of the serial numbers of the parts, and use this sorting as a scheduling sequence.

In the second aspect, a smelting workshop scheduling system based on an improved teaching optimization algorithm is provided, the system includes a memory, a processor, and a computer program stored in the memory and operable on the processor, and the processor executes the The computer program realizes the steps of the above method.

(3) Beneficial effects

The invention provides a smelting workshop scheduling method and system based on an improved teaching optimization algorithm. Compared with the prior art, it has the following beneficial effects:

(1) The present invention proposes an energy efficiency balance optimization model considering the dynamic arrival time of raw materials. In order to solve this model, the original teaching optimization algorithm is innovated, and class initialization, teaching links, adaptive learning links and Bernoulli learning links are respectively carried out. Improvements and innovations in cross-self-learning links and other aspects, combined with the structural characteristics of the model itself, incorporate the above-mentioned local innovation sub-processes into the framework of the improved teaching optimization algorithm. The improved algorithm can describe the teaching links more objectively, self-adaptive In the late stage of the algorithm, the learning steps of the students in the class can be adjusted to improve the self-learning ability of the individual students in the class, and the approximate optimal solution of the above energy efficiency balance optimization model can be solved within a reasonable solution time, providing certain theoretical guidance for the actual production of enterprises with support.

(2) Compared with the relatively simple method of randomly initializing the population in the traditional teaching optimization algorithm, in the improved teaching optimization algorithm, in addition to the traditional method of randomly initializing the population, combined with the characteristics of the energy efficiency balance optimization problem, a waste heat Making full use of the principled initial population quality improvement mechanism can well enhance the search accuracy of the initial population.

(3) The teaching link of the traditional teaching optimization algorithm only involves individual teachers and average students. However, considering the complexity of solving the comprehensive score value, the average student with the average score of each subject cannot be regarded as the average score of the comprehensive score. Individual, therefore, in the improved teaching optimization algorithm, the concept of median student is introduced, which can well express the individual student whose comprehensive grade value is in the middle level.

(4) In order to further reduce the possibility of the algorithm falling into a local optimum, in the improved teaching optimization algorithm, based on the concept of Bernoulli crossover in the biased random secret key algorithm, an innovative self-learning method for students based on Bernoulli crossover is proposed link, which can well balance the depth and breadth of the search.

Description of drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings that need to be used in the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are only These are some embodiments of the present invention. Those skilled in the art can also obtain other drawings based on these drawings without creative work.

Fig. 1 is a flowchart of an embodiment of the present invention.

Detailed ways

In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention are clearly and completely described. Obviously, the described embodiments are part of the embodiments of the present invention, not all of them. example. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without creative efforts fall within the protection scope of the present invention.

The embodiment of the present application provides a smelting workshop scheduling method and system based on an improved teaching optimization algorithm, which solves the problem that the existing scheduling algorithm cannot utilize the waste heat that may exist in the precision investment casting process of high-end equipment, and realizes the oriented Energy-efficient production scheduling.

The technical solution in the embodiment of this application is to solve the above-mentioned technical problems. The general idea is as follows: Considering the energy efficiency balance optimization model of the dynamic arrival time of raw materials, in order to solve this model, the original teaching optimization algorithm is innovated, and the class initialization, Improvements and innovations in teaching, self-adaptive learning, and Bernoulli self-learning, combined with the structural characteristics of the model itself, incorporate the above local innovation sub-processes into the framework of the improved teaching optimization algorithm. The improved algorithm can Describe the teaching process more objectively, adaptively adjust the learning step size of class students in the later stage of the algorithm, improve the self-learning ability of individual class students, and can solve the approximate optimal solution of the above energy efficiency balance optimization model within a reasonable solution time, Provide certain theoretical guidance and support for the actual production of enterprises.

In order to better understand the above-mentioned technical solution, the above-mentioned technical solution will be described in detail below in conjunction with the accompanying drawings and specific implementation methods.

Example 1:

As shown in Figure 1, the present invention provides a smelting workshop scheduling method based on an improved teaching optimization algorithm, which is executed by a computer, and the method includes:

S1. Obtain scheduling information; set the total number of iterations Mit, and initialize the number of iterations it=0;

S2. Randomly initialize the grades of 80% of individual students in each subject in the class, and randomly initialize the grades of each subject of the remaining 20% of individual students based on the principle of fully utilizing the waste heat of the furnace body, and construct the initial grades of each subject of individual students in the class;

S3. Based on the individual students' initial grades in each subject, calculate the comprehensive grades of each individual student in the class, and construct the initial class grades;

S4. Take the individual student with the best overall score in the current class as the individual teacher;

The individual student whose comprehensive score is the average of each subject in the current class is regarded as an equal individual student;

Take the student individual whose comprehensive score is the median in the current class as the median individual student;

S5. Based on the current individual teacher, average student individual and median student individual, carry out the teaching process for the current class, and update the current class grade;

S6. Use the grades of any individual student in the class to perform adaptive learning for the individual students in the class, and update the current grades of the class again;

S7. Execute the self-learning link of individual students in the current class, and update the current class grades to form the class grades after the it-th iteration;

S8. Judging whether the number of iterations it reaches the total number of iterations Mit, if so, output the current optimal student individual score, and convert it into a scheduling sequence through the decoding algorithm; otherwise, update the current individual teacher, equal student individual and median student individual Subject grades, iteration times, and return to S5.

The beneficial effects of this embodiment are:

The present invention proposes an energy efficiency balance optimization model considering the dynamic arrival time of raw materials. In order to solve this model, the original teaching optimization algorithm is innovated, and class initialization, teaching links, adaptive learning links and Bernoulli self-learning links are respectively carried out In combination with the structural characteristics of the model itself, the above local innovation sub-processes are incorporated into the framework of the improved teaching optimization algorithm. The improved algorithm can describe the teaching process more objectively and adaptively in the later stage of the algorithm. Adjusting the learning step size of class students and improving the individual self-learning ability of class students can solve the approximate optimal solution of the above energy efficiency balance optimization model within a reasonable solution time, providing certain theoretical guidance and support for the actual production of enterprises.

The implementation process of the embodiment of the present invention is described in detail below:

S1. Obtain scheduling information; set the total number of iterations Mit, and initialize the number of iterations it=0;

The scheduling information includes:

There are N high-end equipment precision parts that need to be cast in a precision casting workshop with a single melting furnace. The set of high-end equipment parts is recorded as Ω={J ₁ ,…,J _j ,…,J _N }; the serial number of the parts is {1,2,...,N};

The time when the raw materials of all parts arrive at the smelting workshop is inconsistent, but it can be known in advance through the Internet of Things technology, and the collection of dynamic arrival times of parts is recorded as

Since the preheating operation is mainly aimed at the crucible and furnace body system in the furnace, the preheating time is almost the same when processing each component, so the preheating time required for processing each component is the same and recorded as t _p ;

In the smelting and refining stage, because the weight and volume of the raw materials required for each part are different, the required smelting and refining time are different, which are recorded as:

Melting time π={s ₁ ,…,s _j ,…,s _N };

Refining time ω={p ₁ ,...,p _j ,...p _N };

S2. Randomly initialize the scores of 80% of individual students in each subject in the class, and randomly initialize the scores of each subject of the remaining 20% of individual students based on the principle of making full use of the waste heat of the furnace, and construct the initial scores of individual students in each subject in the class, the number of subjects d is the same as the number of parts.

Specifically include the following steps:

S201, randomly generating a random number between [0,1] as the student's individual random score;

S202, 20% of the students' individual random scores are decoded by the decoding algorithm to obtain the corresponding scheduling sequence, which is recorded as

S203. Set j=1, and start the smelting operation of part J _j at the time of max{t _p , r _j }, record the completion time of the refining operation of part J _j as C _j ;

S204. Determine whether the arrival time of the remaining parts does not exceed C _j , if yes, arrange the part with the largest sum of smelting time and refining time among all parts whose arrival time does not exceed C _j to start the smelting operation, and the refining operation is completed within the time Denote as C _j+1 ;

S205, judge whether j is less than N-1, if it is true, set j=j+1 and return to S204; otherwise, use the decoding algorithm to reverse the current scheduling sequence again, and obtain the students corresponding to 20% of the students' individual random scores The individual's initial grades of each subject, together with 80% of the individual random grades of the students, constitute the initial grades of each subject of the individual students in the class.

Wherein, the decoding algorithm is used to convert the achievement of the ith generation of the individual student into a scheduling sequence, and the specific steps are as follows:

Step 1: Combine the scores of individual students in each subject One-to-one correspondence with the part number {1,2,...,N};

Step 2: Align the component serial numbers {1, 2, ..., N} according to the grades of individual students in each subject Sort in ascending order of the size of the new parts and components, and use this sort as the scheduling sequence.

S3. Based on the individual students' initial grades in each subject, calculate the comprehensive grades of each individual student in the class, that is, the energy efficiency balance index of the scheduling sequence; construct the initial class grades;

According to the characteristics of the actual smelting process, it has the following typical characteristics:

① All parts are casted in a non-preemptive processing method, that is, once the casting operation of a certain part is started, the process cannot be interrupted, and the casting of subsequent parts must be completed after the processing of the previous part.

②Assuming that the melting temperature and refining temperature of each part tend to be the same, once the crucible finishes processing the previous part J _j-1 and then processes the next part J _j , the melting time of the latter is αs _j , where α is the waste heat utilization rate, 0<α<1.

③Once the cooling time of the crucible and furnace system exceeds the threshold χ, an additional warm-up time t _p is required for the casting operation of the next component.

④Assume that the preheating power of the crucible and furnace body is P _pre , the melting power is P _s , and the refining power is P _p . The goal of the energy efficiency balance optimization mentioned above is on the one hand to minimize the manufacturing span required to cast all parts, and on the other hand to minimize the total energy consumption of the crucible and furnace system.

Suppose the parts processing sequence is The idle time between two adjacent parts J _j and J _j+1 is Δt _i,i+1 , then the manufacturing span can be calculated as:

Then the total energy consumption of the foundry system can be calculated as:

According to the nature of the problem, the lower bounds of the manufacturing span and total energy consumption of the casting system are respectively:

Combining the above-mentioned manufacturing span and energy consumption lower bound information, the above-mentioned double-objective problem can be transformed into a single-objective problem, and the energy efficiency balance index of the scheduling sequence (that is, the comprehensive score of the individual student) can be expressed as follows:

w _C and w _E are the weights of manufacturing span and total energy consumption, respectively.

Therefore, the specific steps for calculating the comprehensive grades of individual students in the class can be as follows:

S301. Using the decoding algorithm to convert the individual student's initial scores in each subject into a scheduling sequence, denoted as

S302. Set j=1, and start the smelting operation of component J _j at the time max{t _p , r _j }, record the completion time of the refining operation of component J _j as C _j ;

S303. Start the smelting operation of component J _j+1 at the time of max{C _j ,r _j+1 }, record the completion time of the refining operation of component J _j+1 as C _j+1 , and record Δt _j,j+1 = max{ C _j , r _j+1 } - the value of C _j ;

S304. Determine whether j is smaller than N-1. If yes, set j=j+1 and return to step 2; otherwise, calculate and output the energy efficiency balance index, and use it as the comprehensive score of the individual student.

Based on the comprehensive grades of individual students and the grades of each subject, the initial class grades are constructed, which are recorded as:

Among them, f(X ⁱ [0]) represents the initial comprehensive score of the i-th individual student in the class; i=1,...,NP;

X ^NP [0] represents the NPth individual student in the class;

Indicates the initial score of the dth subject of the NPth individual student in the class;

S4. Take the individual student with the best overall score in the current class as the teacher individual X ^best [0], namely:

arg min{f(X ⁱ [0]),i=1,...,NP}

In the current class, the student whose comprehensive score is the average of each subject is taken as the equal student individual X ^avg [0], that is:

Take the individual student whose comprehensive grade is the median in the current class as the median student individual X ^med [0], that is:

S5. Based on the current individual teacher, average student individual and median student individual, carry out the teaching process for the current class, and update the current class grade;

The teaching process mainly uses the grades of individual teachers, average students, and median students to iterate the current class. The specific steps are as follows:

S501. Initialize i=1, it=0; that is, when performing the first iteration, iterate based on the initial grades of the class.

The grades and comprehensive grades of individual students in the ith generation class are specifically expressed as follows:

Individual teachers' grades in each subject and comprehensive grades are recorded as:

The scores of each subject and the comprehensive score of the median individual student are recorded as:

The scores of each subject and the comprehensive score of each individual student are recorded as:

S502, let j=1;

S503. Let rand _i =rand(0,1), TF _i =1+rand(0,1),

And according to the following formula to construct the first class of new student individual subject j grades;

S504. Let rand _i =rand(0,1) and TF _i =1+rand(0,1),

And according to the following formula to construct the second type of new individual students' grades of subject j;

S505. Determine whether j≤N is established, if established, then set j=j+1 and turn to S503; otherwise, calculate the comprehensive scores of the two new types of student individuals New1X ⁱ [it] and New2X ⁱ [it] respectively, and Keep the one with the best comprehensive score among X ⁱ [it], New1X ⁱ [it] and New2X ⁱ [it] as X ⁱ [it+1];

S506. Determine whether i≤NP holds true, if true, set i=i+1, update individual teachers, average students, and median students, and turn to S502; otherwise, the teaching link is terminated.

S6. Use the grades of any individual student in the class to perform adaptive learning for the individual students in the class, and update the current grades of the class again;

The adaptive link mainly simulates the mutual learning of any two individual students in the class, and the learning step increases with the increase of the number of iterations, which is convenient for increasing the diversity of class grades in the late iteration. Specific steps are as follows:

S601, setting i=1,

S602. Select any individual student who is different from Xi [it] as X ^k [it], and ^set j=1;

S603. Order Where θ>1, and construct a new individual student's grade j subject according to the following formula;

S604. Determine whether j≤N is true, if true, set j=j+1, and turn to S603; otherwise, calculate the comprehensive score of the new student individual NewX ⁱ [it], and compare ^Xi [it] and NewX The one with the best overall score in ⁱ [it] is reserved as X ⁱ [it+1];

S605. Determine whether i≤NP holds true, if true, set i=i+1, and go to S602; otherwise, the adaptive learning link is terminated.

S7. Execute the self-learning link of individual students in the current class, and update the current class grades to form the class grades after the it-th iteration;

The Bernoulli self-study link mainly simulates the self-study situation of a single student in the class. The self-renewal of the individual student's grades in each subject is carried out through the Bernoulli cross-variation method. The goal of the update is to achieve the individual teacher's performance in each subject score. Specific steps are as follows:

S701: set i=1,

S702, setting j=1;

S703. Let BR _i =rand(0,1), and construct the grade of subject j of a new individual student according to the following formula;

S704. Determine whether j≤N is true, if true, set j=j+1 and go to S703; otherwise, calculate the comprehensive score of the new student individual NewBX ⁱ [it], and compare ^Xi [it] and NewBX ⁱ The one with the best comprehensive score in [it] is reserved as X ⁱ [it+1];

S705: Determine whether i≤NP holds true, if true, set i=i+1, and go to S702; otherwise, the Bernoulli self-learning learning link is terminated.

S8. Judging whether the number of iterations it reaches the total number of iterations Mit, if so, output the current optimal student individual score, and convert it into a scheduling sequence through the decoding algorithm; otherwise, update the current individual teacher, equal student individual and median student individual Subject grades, iteration times, and return to S5.

Example 2

The present invention also provides a smelting workshop scheduling system based on the improved teaching optimization algorithm, the system includes a memory, a processor, and a computer program stored in the memory and operable on the processor, and the processor executes the computer program The procedure is to realize the steps of the above-mentioned method.

It can be understood that the smelting workshop scheduling system based on the improved teaching optimization algorithm provided by the embodiment of the present invention corresponds to the above-mentioned smelting workshop scheduling method based on the improved teaching optimization algorithm. The corresponding content in the smelting shop scheduling method based on the improved teaching optimization algorithm will not be repeated here.

In summary, compared with the prior art, the present invention has the following beneficial effects:

The present invention proposes an energy efficiency balance optimization model considering the dynamic arrival time of raw materials. In order to solve this model, the original teaching optimization algorithm is innovated, and class initialization, teaching links, adaptive learning links and Bernoulli self-learning links are respectively carried out In combination with the structural characteristics of the model itself, the above local innovation sub-processes are incorporated into the framework of the improved teaching optimization algorithm. The improved algorithm can describe the teaching process more objectively and adaptively in the later stage of the algorithm. Adjusting the learning step size of class students and improving the individual self-learning ability of class students can solve the approximate optimal solution of the above energy efficiency balance optimization model within a reasonable solution time, providing certain theoretical guidance and support for the actual production of enterprises.

It should be noted that, through the above description of the implementation manners, those skilled in the art can clearly understand that each implementation manner can be implemented by means of software plus a necessary general hardware platform. Based on this understanding, the essence of the above technical solution or the part that contributes to the prior art can be embodied in the form of software products, and the computer software products can be stored in computer-readable storage media, such as ROM/RAM, magnetic discs, optical discs, etc., including several instructions to make a computer device (which may be a personal computer, server, or network device, etc.) execute the methods described in various embodiments or some parts of the embodiments. In this document, relational terms such as first and second etc. are used only to distinguish one entity or operation from another without necessarily requiring or implying any such relationship between these entities or operations. Actual relationship or sequence. Furthermore, the term "comprises", "comprises" or any other variation thereof is intended to cover a non-exclusive inclusion such that a process, method, article or apparatus comprising a set of elements includes not only those elements, but also includes elements not expressly listed. other elements of or also include elements inherent in such a process, method, article, or device. Without further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of additional identical elements in the process, method, article or apparatus comprising said element.

The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: it can still be described in the foregoing embodiments Modifications are made to the recorded technical solutions, or equivalent replacements are made to some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims (9)

1. A smelting workshop scheduling method based on an improved teaching optimization algorithm, characterized in that the method comprises: S1. Obtain scheduling information; set the total number of iterations, and initialize the number of iterations it=0; S2. Randomly initialize the grades of 80% of individual students in each subject in the class, and randomly initialize the grades of each subject of the remaining 20% of individual students based on the principle of fully utilizing the waste heat of the furnace, and construct the initial grades of each subject of individual students in the class. The quantity is the same as the quantity of parts; S3. Based on the individual students' initial grades in each subject, calculate the comprehensive grades of each individual student in the class, and construct the initial class grades; S4. Take the individual student with the best overall score in the current class as the individual teacher; The individual student whose comprehensive score is the average of each subject in the current class is regarded as an equal individual student; Take the student individual whose comprehensive score is the median in the current class as the median individual student; S5. Based on the current individual teacher, average student individual and median student individual, carry out the teaching process for the current class, and update the current class grade; S6. Use the grades of any individual student in the class to perform adaptive learning for the individual students in the class, and update the current grades of the class again; S7. Execute the self-learning link of individual students in the current class, and update the current class grades to form the class grades after the it-th iteration; S8. Judging whether the number of iterations it reaches the total number of iterations Mit, if so, output the current optimal student individual score, and convert it into a scheduling sequence through the decoding algorithm; otherwise, update the current individual teacher, equal student individual and median student individual Subject grades, iteration times, and return to S5. 2. a kind of smelting workshop scheduling method based on improved teaching optimization algorithm as claimed in claim 1, is characterized in that, described scheduling information comprises: Parts set Ω＝{J ₁ ,…,J _j ,…,J _N }; A collection of component dynamic arrival times The warm-up time t _p required by the parts; Melting time π={s ₁ ,…,s _j ,…,s _N }; Refining time ω={p ₁ ,...,p _j ,...p _N }. 3. A method for scheduling smelting workshops based on an improved teaching optimization algorithm as claimed in claim 1, wherein said S2, randomly initializing the grades of 80% of the individual students in the class, is based on the full utilization of waste heat in the furnace body The principle is to randomly initialize the grades of each subject of the remaining 20% of the individual students, and the specific steps to construct the initial grades of each subject of the individual students in the class are as follows: S201, randomly generating a random number between [0,1] as the student's individual random score; S202, 20% of the students' individual random scores are decoded by the decoding algorithm to obtain the corresponding scheduling sequence, which is recorded as N is the number of parts; S203. Set j=1, and start the smelting operation of part J _j at the time of max{t _p , r _j }, record the completion time of the refining operation of part J _j as C _j ; t _p represents the warm-up time required by the component, and r _j represents the dynamic arrival time of the jth component; S204. Determine whether the arrival time of the remaining parts does not exceed C _j , if yes, arrange the part with the largest sum of smelting time and refining time among all parts whose arrival time does not exceed C _j to start the smelting operation, and the refining operation is completed within the time Denote as C _j+1 ; S205, judge whether j is less than N-1, if it is true, set j=j+1 and return to S204; otherwise, use the decoding algorithm to reverse the current scheduling sequence again, and obtain the students corresponding to 20% of the students' individual random scores The individual's initial grades of each subject, together with 80% of the individual random grades of the students, constitute the initial grades of each subject of the individual students in the class. 4. A kind of smelting workshop scheduling method based on improved teaching optimization algorithm as claimed in claim 3, is characterized in that, S5, described based on current teacher individual, equal student individual and median student individual are carried out to current class The teaching link, and update the current class grades, specifically include the following steps: S501. Initialize i=1, it=0; that is, when performing the first iteration, iterate based on the initial class grade; The grades and comprehensive grades of the individual students in the it-th generation class are: Among them, f(X ⁱ [it]) represents the initial comprehensive score of the i-th individual student in the class of the it-th generation; i=1,...,NP; X ^NP [it] represents the NP-th student individual in the class of the it-th generation; Indicates the initial score of the d-th subject of the NP-th individual student in the class of the it-th generation; The grades and comprehensive grades of individual teachers in each subject are as follows: The median grades of each subject and the comprehensive score of individual students are: The scores of each subject and the comprehensive score of the average student are: S502, let j=1; S503. Let rand _i =rand(0,1), TF _i =1+rand(0,1), And construct the grades of the first class of new student individual subject j: S504. Let rand _i =rand(0,1) and TF _i =1+rand(0,1), And construct the grades of the jth subject of the second type of new student individual; the construction formula is: S505. Determine whether j≤N is established, if established, then set j=j+1 and turn to S503; otherwise, calculate the comprehensive scores of the two new types of student individuals New1X ⁱ [it] and New2X ⁱ [it] respectively, and Keep the one with the best comprehensive score among X ⁱ [it], New1X ⁱ [it] and New2X ⁱ [it] as X ⁱ [it+1]; S506. Determine whether i≤NP holds true, if true, set i=i+1, update individual teachers, average students, and median students, and turn to S502; otherwise, the teaching link is terminated. 5. A kind of smelting workshop scheduling method based on improved teaching optimization algorithm as claimed in claim 4, it is characterized in that, described S6, utilizes the grade of each subject of any individual student in the class to carry out self-adaptive learning for the individual student in the class link, and update the current class grade again, including the following steps: S601, set i=1; S602. Select any individual student who is different from Xi [it] as X ^k [it], and ^set j=1; S603. Order Where θ>1, and construct the grade of the jth subject of the new individual student: S604. Determine whether j≤N is true, if true, set j=j+1, and turn to S603; otherwise, calculate the comprehensive score of the new student individual NewX ⁱ [it], and compare ^Xi [it] and NewX The one with the best overall score in ⁱ [it] is reserved as X ⁱ [it+1]; S605. Determine whether i≤NP holds true, if true, set i=i+1, and go to S602; otherwise, the adaptive learning link is terminated. 6. a kind of smelting workshop scheduling method based on improved teaching optimization algorithm as claimed in claim 5, it is characterized in that, described S7, carry out the self-study link of individual student in the current class, and update current class result, form the ith Class grades after the second iteration, the specific steps are: S701: set i=1, S702, setting j=1; S703. Let BR _i =rand(0,1), and construct the grade of subject j of a new individual student: S704. Determine whether j≤N is true, if true, set j=j+1 and go to S703; otherwise, calculate the comprehensive score of the new student individual NewBX ⁱ [it], and compare ^Xi [it] and NewBX ⁱ The one with the best comprehensive score in [it] is reserved as X ⁱ [it+1]; S705: Determine whether i≤NP holds true, if true, set i=i+1, and go to S702; otherwise, the Bernoulli self-learning learning link is terminated. 7. A method for scheduling smelting workshops based on an improved teaching optimization algorithm as described in any one of claims 1-6, wherein the formula for calculating the overall performance of the individual student is: Among them, C _max represents the manufacturing span of the casting system, ∑E represents the total energy consumption of the casting system, the lower bounds of the manufacturing span and total energy consumption of the LB _C and LB _E casting systems, w _C and w _E are the manufacturing span and total energy consumption, respectively the weight of energy; Among them, s _j represents the melting time of the jth component; p _j represents the refining time of the jth component; α is the utilization rate of waste heat, which means 0<α<1; Δt _{j, j+1} represents the interval time between two parts processing; t _p represents the warm-up time required by the parts χ represents the cooling timeout threshold; P _s represents the smelting power; P _p represents the refining power; P _pre represents the preheating power. 8. A kind of smelting workshop scheduling method based on the improved teaching optimization algorithm as described in any one of claims 1-6, it is characterized in that, described decoding algorithm is used for converting the achievement of student individual it generation into scheduling sequence, The specific steps are: Step 1: Make a one-to-one correspondence between the grades of individual students in each subject and the serial numbers of parts; Step 2: Sort the serial numbers of the parts in ascending order according to the grades of individual students in each subject to obtain a new sorting of the serial numbers of the parts, and use this sorting as a scheduling sequence. 9. A smelting workshop scheduling system based on an improved teaching optimization algorithm, the system includes a memory, a processor, and a computer program stored on the memory and operable on the processor, wherein the processor executes the The computer program realizes the steps of the method described in any one of claims 1 to 8 above.