Multi-AGV green job shop integrated scheduling method and system based on improved genetic algorithm

Abstract

The invention discloses a multi-AGV green job shop integrated scheduling method and system based on an improved genetic algorithm, and the method comprises the steps: S1, constructing a multi-AGV green job shop scheduling problem model, wherein the model takes minimization of total energy consumption, maximum completion time and total cost as optimization targets, and comprises energy consumption constraints and charging constraints of the AGV in waiting, no-load and full-load states and energy consumption constraints of processing equipment; s2, solving the scheduling problem model by adopting an improved genetic algorithm; wherein chromosome coding of the improved genetic algorithm comprises process sequence coding, equipment allocation coding and AGV allocation coding; and S3, outputting an optimal scheduling scheme through the improved genetic algorithm, and scheduling the AGV and the processing equipment in the workshop according to the output optimal scheduling scheme.

Description

Technical Field

[0001] This invention relates to the field of production system optimization technology, and in particular to a multi-AGV green workshop integrated scheduling method and system based on an improved genetic algorithm. Background Technology

[0002] The deepening advancement of intelligent manufacturing is driving the global manufacturing industry to accelerate its transformation towards intelligence and green practices. As the physical embodiment of smart factories, the construction of smart workshops has become the core of achieving this transformation. In this process, Automated Guided Vehicles (AGVs), due to their advantages of high flexibility, high efficiency, and high safety, are increasingly being used in workshops to handle material handling tasks between warehouses and processing equipment. The Flexible Job Shop Scheduling Problem (FJSP), as a core problem in workshop scheduling, has always received considerable attention from academia and industry. Solving the FJSP and achieving integrated scheduling of AGVs and processing equipment in a multi-AGV collaborative handling environment is a crucial foundation for building green factories.

[0003] In actual production, workshop energy consumption mainly stems from the operation of processing equipment and AGVs, and the energy consumption characteristics of equipment vary significantly under different conditions. Therefore, researching the integrated scheduling problem of green workshops based on multi-AGV transportation is of great significance for implementing the national green manufacturing concept and promoting sustainable industrial development. The workshop scheduling problem is a typical NP-hard problem, currently solved mainly by intelligent optimization algorithms. Researchers have proposed and improved various optimization methods for different problem scenarios. Among them, the Genetic Algorithm (GA), as one of the earliest algorithms introduced into this field, has seen its basic and improved versions widely applied and validated in workshop scheduling.

[0004] However, existing research and technologies still have the following shortcomings: First, there is a lack of systematic research on the integrated scheduling problem of multiple AGVs in green workshops, and there is a lack of research that considers the energy consumption of AGVs and processing equipment under different states, while this problem has important engineering significance in practical applications; Second, traditional intelligent optimization algorithms may encounter problems of insufficient convergence and solution quality when dealing with the integrated scheduling model of AGVs and processing equipment due to the NP-hard nature of the problem.

[0005] Therefore, in view of the above-mentioned prior art, the present invention provides a multi-AGV green workshop integrated scheduling method and system based on an improved genetic algorithm. Summary of the Invention

[0006] The purpose of this invention is to address the shortcomings of existing technologies by providing an integrated scheduling method and system for multi-AGV green workshops based on an improved genetic algorithm. This method can construct a model that simultaneously considers the energy consumption of AGVs and processing equipment under different states, and performs crossover and mutation on chromosomes based on AGVs and equipment. This method can adapt to the multi-constraint characteristics of green workshops and improve the solution effect of the algorithm.

[0007] To achieve the above objectives, the present invention adopts the following technical solution:

[0008] An integrated scheduling method for multi-AGV green workshops based on an improved genetic algorithm includes:

[0009] S1. Construct a scheduling model for a green workshop with multiple AGVs. The model aims to minimize total energy consumption, maximum completion time, and total cost. It also includes energy consumption constraints, charging constraints, and energy consumption constraints of processing equipment for AGVs in waiting, empty, and fully loaded states.

[0010] S2. An improved genetic algorithm is used to solve the scheduling problem model; the chromosome encoding of the improved genetic algorithm includes process sequence encoding, equipment allocation encoding, and AGV allocation encoding.

[0011] S3. Improve the genetic algorithm to output the optimal scheduling scheme, and schedule the AGVs and processing equipment in the workshop according to the output optimal scheduling scheme.

[0012] Furthermore, the objective functions for total energy consumption, maximum completion time, and total cost in step S1 are respectively expressed as:

[0013]

[0014]

[0015]

[0016] Where f1 represents the objective function of total energy consumption; f2 represents the objective function of maximum completion time; f3 represents the objective function of total cost; E operation E represents the energy consumption during equipment processing; idle E represents the energy consumption of the equipment when it is not under load. agv-wait E represents the energy consumption of the AGV while it is waiting; agv-idle E represents the energy consumption of the AGV when it is unloaded; agv-load Indicates the energy consumption of the AGV when it is fully loaded; C i Indicates the completion time of the workpiece; R represents the number of AGVs; r represents the current r-th AGV; C IJr Indicates the time when the AGV stops transporting the workpiece; C AGV Indicates; C shopThis represents the cost incurred by the workshop per unit of time.

[0017] Furthermore, the procedure before step S2 includes:

[0018] The scheduling problem of multi-AGV green workshop is transformed into a single-objective optimization problem, and a comprehensive objective function is constructed.

[0019] Furthermore, the charging constraints of the AGV include: when the AGV's power is below a threshold, if it is in a waiting or empty state, it will go directly to the charging station to charge; if it is in a fully loaded state, it will complete the current handling task first and then go to the charging station to charge.

[0020] Furthermore, in step S2, the improved genetic algorithm uses a load balancing mechanism to allocate processing equipment and AGVs to the process when initializing the population, so as to minimize the cumulative task load variance of all processing equipment and all AGVs.

[0021] Furthermore, in step S2, the selection operation of the improved genetic algorithm adopts a combination of roulette wheel selection and elite retention strategy, directly retaining a predetermined proportion of individuals with the best fitness in the population to the next population.

[0022] Furthermore, in step S2, the crossover operation of the improved genetic algorithm adopts an adaptive crossover probability. The adaptive crossover probability uses a non-linear decreasing strategy according to the iteration process, expressed as:

[0023]

[0024] Among them, P c P represents the crossover probability; cmin P represents the minimum crossover probability; cmax Gen represents the maximum crossover probability; Gen represents the current iteration number; GenMax represents the maximum iteration number; k represents the decay coefficient.

[0025] Furthermore, the crossover operation specifically involves: randomly selecting a target workpiece from the parent chromosomes, and for at least two discontinuous processes of the selected target workpiece, exchanging the equipment allocation gene segment and AGV allocation gene segment from the two parent chromosomes corresponding to the current process.

[0026] Furthermore, the mutation operation of the improved genetic algorithm in step S2 includes: randomly selecting a workpiece in the chromosome and exchanging the equipment allocation codes corresponding to any two processes within the selected workpiece; randomly selecting two gene positions in the AGV coding segment of the chromosome and exchanging the AGV numbers.

[0027] Correspondingly, an integrated scheduling system for multi-AGV green workshops based on an improved genetic algorithm is also provided, including:

[0028] The building module is used to construct a scheduling problem model for a multi-AGV green workshop. The model aims to minimize total energy consumption, maximum completion time and total cost, and includes energy consumption constraints, charging constraints and energy consumption constraints of processing equipment for AGVs in waiting, empty and full-load states.

[0029] The solution module is used to solve the scheduling problem model using an improved genetic algorithm; the chromosome encoding of the improved genetic algorithm includes process sequence encoding, equipment allocation encoding, and AGV allocation encoding.

[0030] The output module is used to output the optimal scheduling scheme through an improved genetic algorithm, and to schedule the AGVs and processing equipment in the workshop according to the output optimal scheduling scheme.

[0031] Compared with the prior art, the present invention has the following beneficial effects:

[0032] 1. A mathematical model was constructed with the objectives of minimizing energy consumption, maximizing completion time, and minimizing cost in the workshop, and the above multiple objectives were unified into a single objective for optimization. The model considers the energy consumption and operation time of AGVs in three states: idle, fully loaded, and waiting, and introduces charging constraints for AGVs.

[0033] 2. An IGA was designed for model solving. A load balancing mechanism was introduced in the initialization phase to distribute the processes to AGVs and processing equipment as evenly as possible. The crossover probability was adaptively adjusted, and the algorithm was improved based on the processes, AGVs, and processing equipment. Attached Figure Description

[0034] Figure 1 Here is a flowchart of the multi-AGV green workshop integrated scheduling method based on an improved genetic algorithm provided in Example 1;

[0035] Figure 2 This is a flowchart of the improved genetic algorithm provided in Implementation Example 1;

[0036] Figure 3 This is an example of a chromosome provided in Example 1;

[0037] Figure 4 This is a schematic diagram of chromosome crossing over provided in Example 1;

[0038] Figure 5 These are iteration diagrams of MK01 examples under different algorithms provided in Example 2;

[0039] Figure 6 These are iteration diagrams of MK02 examples under different algorithms provided in Example 2;

[0040] Figure 7This refers to the percentage of working states under different numbers of AGVs in the configuration provided in Example 2;

[0041] Figure 8 These are different target values ​​for different numbers of AGVs configured in Example 2. Detailed Implementation

[0042] The following specific examples illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that, unless otherwise specified, the following embodiments and features described therein can be combined with each other.

[0043] The purpose of this invention is to address the shortcomings of existing technologies by providing an integrated scheduling method and system for multi-AGV green work workshops based on an improved genetic algorithm.

[0044] Example 1

[0045] This embodiment provides an integrated scheduling method for multi-AGV green workshops based on an improved genetic algorithm, such as... Figure 1-2 As shown, it includes:

[0046] S1. Construct a scheduling model for a green workshop with multiple AGVs. The model aims to minimize total energy consumption, maximum completion time, and total cost. It also includes energy consumption constraints, charging constraints, and energy consumption constraints of processing equipment for AGVs in waiting, empty, and fully loaded states.

[0047] S2. An improved genetic algorithm is used to solve the scheduling problem model; the chromosome encoding of the improved genetic algorithm includes process sequence encoding, equipment allocation encoding, and AGV allocation encoding.

[0048] S3. Improve the genetic algorithm to output the optimal scheduling scheme, and schedule the AGVs and processing equipment in the workshop according to the output optimal scheduling scheme.

[0049] This embodiment uses a machining workshop as an example. The workshop layout includes a material storage area S. t A finished product area E n and a charging station C h .

[0050] The scheduling optimization problem of a multi-AGV green operation workshop can be simply described as follows: There are N workpieces in the material area of ​​the workshop ( ) in M ​​devices ( The AGVs are processed on the finished product area and then placed in the finished product area. There are a total of R AGVs. ) is used for workpiece handling, each workpiece contains J processes ( The number of processes is no greater than the number of equipment units. Each process can only be processed on one piece of equipment. Any AGV can be selected for the handling of the workpiece.

[0051] AGVs handle three scenarios during transport: waiting, idle, and fully loaded. When an AGV's battery level drops below a threshold, it needs to recharge at a charging station. If the AGV is waiting or idle, it goes directly to the charging station. If the AGV is fully loaded, it must first move the workpiece to the target location before going to the charging station. The cost incurred by an AGV per unit time is... The cost incurred by the workshop per unit time is The j-th process for workpiece i is... The completion time of workpiece i is , The initial handling time on the AGV r is , The termination time of the transport on the AGV r is AGV r transport The previous process The termination time is , The initial processing time on device m is , The termination time on device m is Equipment m processing The previous process The completion time is , The processing time on device m is The energy consumption per unit time during equipment processing is The energy consumption per unit time when the equipment is idle is The energy consumption per unit time when the AGV is waiting is The energy consumption of an AGV per unit time when it is unloaded is The energy consumption per unit time when the AGV is fully loaded is The maximum power of AGV r is The actual power consumption is Energy consumption per unit time is Charging time is The charging rate is , for At the available time of process j for workpiece i. For the j-process of workpiece i The delivery time Let m be the processing time for operation j of workpiece i. Let be the completion time of process j for workpiece i. For workpiece Process The completion time, for From the process The time it takes for the device m to reach the charging station. for From the process The time it takes for the device m to reach the charging station. For 0-1 variables, process The value is 1 if the processing is performed by equipment m, and 0 otherwise. For 0-1 variables, process The value is 1 if the AGV r is used for transportation, and 0 otherwise. For 0-1 variables, AGV r handling process If charging is required, the value is 1; otherwise, it is 0.

[0052] In step S1, a multi-AGV green workshop scheduling problem model is constructed. The model aims to minimize total energy consumption, maximum completion time, and total cost. It also includes energy consumption constraints, charging constraints, and energy consumption constraints of processing equipment for AGVs in waiting, empty, and fully loaded states.

[0053] S11. Construct three major optimization objective functions;

[0054] Objective 1: Minimize total energy consumption f1; where total energy consumption is the sum of equipment energy consumption and AGV energy consumption, expressed as:

[0055]

[0056] Among them, E operation E represents the energy consumption during equipment processing; idle E represents the energy consumption of the equipment when it is not under load. agv-wait E represents the energy consumption of the AGV while it is waiting; agv-idle E represents the energy consumption of the AGV when it is unloaded; agv-load This indicates the energy consumption of the AGV when it is fully loaded.

[0057] Objective 2: Minimize the maximum completion time f2; where the maximum completion time C max It is a key indicator for measuring production efficiency.

[0058]

[0059] Among them, C i Indicates the completion time of the workpiece.

[0060] Objective 3: Minimize the total cost f3; where the total cost includes variable costs (AGV usage costs) and fixed costs (shop time costs).

[0061]

[0062] Where R represents the number of AGVs; r represents the current r-th AGV; C IJr Indicates the time when the AGV stops transporting the workpiece; C AGV Indicates; C shop This represents the cost incurred by the workshop per unit of time.

[0063] S12. Determine the constraints of the problem model, including:

[0064]

[0065]

[0066]

[0067]

[0068]

[0069]

[0070]

[0071]

[0072]

[0073]

[0074]

[0075]

[0076]

[0077]

[0078]

[0079]

[0080]

[0081]

[0082]

[0083] Wherein, equation (4) represents the energy consumption of the equipment during processing; equation (5) represents the energy consumption of the equipment when it is unloaded;

[0084] Equation (6) represents the energy consumption of the AGV while waiting; Equation (7) represents the energy consumption of the AGV when it is idle; Equation (8) represents the energy consumption of the AGV when it is fully loaded; Equation (9) represents the completion time of the workpiece when the last process of processing is completed; Equation (10) represents the process time when the AGV starts to transport the workpiece, which is constrained by the AGV idle time, the AGV transport time, and the completion time of the previous process; Equation (11) represents the idle time of the AGV, which is not earlier than the end time of the previous transport task; Equation (12) represents the time when the AGV stops transporting the workpiece, which is constrained by the start time of transport and the AGV transport time. Constraints; Equation (13) indicates that the start time of the workpiece's process is constrained by the time the AGV finishes transporting the process and the completion time of the previous process processed by the equipment processing the process; Equation (14) indicates that the completion time of the workpiece's process is constrained by the start time of the process and the processing time of the process on the equipment; Equations (15) to (17) all indicate the total transportation time for the AGV to complete the task; Equations (18) to (20) respectively indicate the waiting time, full load time and idle time of the AGV; Equation (21) indicates the actual electricity consumed by the AGV; Equation (22) indicates the charging time of the AGV.

[0085] In this embodiment, the charging constraints of the AGV include: when the AGV's power is below a threshold, if it is in a waiting or empty state, it will go directly to the charging station to charge; if it is in a fully loaded state, it will complete the current handling task first and then go to the charging station to charge.

[0086] S13. The multi-AGV green operation workshop scheduling problem model is merged into a single-objective optimization problem through a weighted normalization method, and a comprehensive objective function f is constructed.

[0087] Through three independent fast optimizations, the theoretical optimal values ​​for the three single objectives are obtained respectively. Then, a comprehensive objective function is constructed, expressed as:

[0088]

[0089] Among them, F1 represents the goal of obtaining the minimum energy consumption, F2 represents the goal of obtaining the minimum maximum completion time, and F3 represents the goal of obtaining the minimum cost.

[0090] In step S2, an improved genetic algorithm is used to solve the scheduling problem model; the chromosome encoding of the improved genetic algorithm includes process sequence encoding, equipment allocation encoding, and AGV allocation encoding.

[0091] S20. Define the current iteration count Gen, the maximum iteration count GenMax, the population size Nind, and the maximum crossover probability. Minimum crossover probability Probability of mutation The fitness function is fitness, and the target value is assigned to fitness.

[0092] The chromosome is defined by encoding processes, processing equipment, and AGVs. Chromosomes use integer encoding, with a length equal to the total number of processes plus the total number of workpieces. Figure 3 Taking chromosome encoding as an example, a three-segment integer encoding is used for 3 workpieces (each with 2 processes):

[0093] Process coding: A permutation based on workpiece numbers. For example, the coding [2, 1, 3, 3, 1, 2, 3, 2, 1] indicates a scheduling sequence of: the first process of workpiece 2, the first process of workpiece 1, the first process of workpiece 3, the second process of workpiece 3, the second process of workpiece 1, the second process of workpiece 2, workpiece 3 is processed and sent to the finished product area, workpiece 2 is processed and sent to the finished product area, and workpiece 1 is processed and sent to the finished product area. Here, 1, 2, and 3 all represent workpiece numbers, and the kth occurrence represents the kth process of that workpiece; this sequence determines the processing and handling order of all processes.

[0094] Processing equipment: The length is the same as the process code, for example [2, 1, 2, 1, 1, 2, 0, 0, 0], where 1 and 2 represent the processing equipment numbers selected for each process, and 0 is a special code indicating that the workpiece has been processed in the previous process and is now sent to the finished product area.

[0095] AGV code: The code is the same length as the process code, for example, [1, 1, 2, 1, 2, 1, 2, 2, 1]. 1 and 2 represent the AGV numbers that perform the corresponding process handling tasks. For example, the first number 1 indicates that workpiece 2 is handled by AGV 1 from the material area to the first process.

[0096] S21. Population initialization.

[0097] When initializing the population, a load balancing mechanism is introduced. Specifically, when randomly assigning equipment and AGVs to processes, the equipment and AGVs with the shortest current cumulative task time are selected first, so as to minimize the cumulative task load variance of all processing equipment and all AGVs, thereby generating a batch of relatively balanced initial solutions and accelerating the convergence of the algorithm.

[0098] S22. Calculate the fitness value based on the chromosome and the corresponding objective function.

[0099] For each individual (chromosome) in the population, based on the scheduling scheme represented by its encoding, f1, f2, and f3 are optimized individually to obtain three theoretically optimal reference values: f 1min , f 2min , f 3min .

[0100] Calculate the overall fitness: For each chromosome (scheduling scheme) in the population, calculate its corresponding f1, f2, f3 values ​​by combining equations (1), (2), and (3), and then substitute them into equation (23) to calculate its overall fitness f; where the smaller the f value, the better the overall performance of the scheme.

[0101] S23. A combination of roulette wheel selection and elite preservation is used to select chromosomes.

[0102] By directly copying the top 5% of individuals (elites) with the highest overall fitness f in the current population into the next generation, the algorithm can ensure that it does not lose the optimal solution it has already discovered.

[0103] For the remaining 95% of positions in the next generation population, a roulette wheel method is used to select parent individuals from the current entire population; the smaller the individual's fitness value f, the greater its probability of being selected.

[0104] S24. The two chromosomes of the father produce new chromosomes in the offspring by exchanging genes with each other.

[0105] Crossover is the main way to generate new individuals. This embodiment uses adaptive crossover probability and segment crossover for process steps.

[0106] The adaptive crossover probability adopts a nonlinear decreasing strategy based on the iteration process, resulting in faster convergence in the early stages of chromosome convergence. As the number of iterations increases, the convergence speed slows down, as expressed in:

[0107]

[0108] Among them, P c P represents the crossover probability; cmin P represents the minimum crossover probability; cmaxGen represents the maximum crossover probability; Gen represents the current iteration number; GenMax represents the maximum iteration number; k represents the decay coefficient.

[0109] Since the available equipment for the corresponding process has been determined, and the equipment number for the last process of the workpiece is 0, only one process is randomly selected to complete the chromosome crossover operation. For example... Figure 2 The chromosome shown contains processes for 3 workpieces. A random number 2 (from 1 to 3) is generated. Then, a crossover operation is performed on the processes for workpiece 2 using PMX. Two random numbers, 3 and 5, representing the number of processes for workpiece 2 are selected. At this point, the final process of transporting the finished product to the warehouse is not selected. The equipment and AGV codes for these two processes are then crossovered. Since the same process can use the same equipment for processing and different AGVs for transport, these two genes can be directly crossovered. The crossover method is as follows: Figure 4 As shown.

[0110] S25. Adopt a mutation operation strategy based on random two-point swapping.

[0111] With a fixed small probability P m =0.1 to mutate individuals to avoid the algorithm getting trapped in local optima.

[0112] Since each workpiece has a designated processing equipment, but the workpiece can be transported by any AGV, when performing mutation operations on the chromosome, a workpiece and the number of processes it has can be randomly selected. The processing equipment at the positions of these two numbers within the workpiece is crossed, and then two numbers not greater than the code number are randomly generated. The AGV codes at these two positions are crossed to obtain a new offspring chromosome.

[0113] This two-dimensional variation in the present embodiment can simultaneously change the processing path and the transport path, thereby enhancing the diversity of the search.

[0114] S26. Iteration and Termination.

[0115] Determine if Gen ≤ GenMax. If yes, use the new population generated after selection, crossover, and mutation as the current population, and return to step S22 for the next round of fitness evaluation and evolutionary operations (Gen = Gen + 1). Repeat this process until the preset maximum number of iterations GenMax (e.g., 200 generations) is reached, at which point the algorithm terminates and outputs the individual with the highest fitness among all generations of the population.

[0116] In step S3, the optimal scheduling scheme is output by improving the genetic algorithm, and the AGVs and processing equipment in the workshop are scheduled according to the output optimal scheduling scheme.

[0117] When the algorithm terminates, the chromosome with the smallest overall fitness f-value is selected from the final population as the optimal scheduling scheme. This chromosome is then decoded to generate an executable scheduling instruction table, including:

[0118] For each workpiece and each process, the start / end transportation time, the AGV used, the start / end processing time, and the equipment used; the transportation task sequence, charging time and duration for each AGV; and the processing task sequence for each piece of equipment.

[0119] Based on this scheduling scheme, the workshop control system sends control commands to the corresponding AGVs and equipment to execute production tasks.

[0120] In this embodiment, S4 is also included: optimizing and analyzing the AGV configuration.

[0121] This embodiment can also be used to assist workshops in making investment decisions, namely, determining the optimal number of AGVs.

[0122] Under the same workshop model and task set, the number of AGVs is set to R units.

[0123] For each configuration quantity, the AGV scheduling method of this embodiment is run to obtain its optimal scheduling scheme, and key performance indicators under this scheme are recorded, such as: total energy consumption f1, maximum completion time C. max The total cost f3, as well as the percentage of AGVs working at full capacity, working at idle capacity, and waiting capacity.

[0124] If the goal is to achieve high equipment utilization and low cost, one might choose to wait for the inflection point before the proportion begins to rise significantly.

[0125] If the ultimate goal is production efficiency (shortest completion time), then adding AGVs can shorten the C... max The quantity where the effect is no longer obvious.

[0126] Managers can use this information to weigh efficiency, energy consumption, and cost, and make informed resource allocation decisions.

[0127] Compared with the prior art, this embodiment has the following beneficial effects:

[0128] 1. A mathematical model was constructed with the objectives of minimizing energy consumption, maximizing completion time, and minimizing cost in the workshop, and the above multiple objectives were unified into a single objective for optimization. The model considers the energy consumption and operation time of AGVs in three states: idle, fully loaded, and waiting, and introduces charging constraints for AGVs.

[0129] 2. An IGA was designed for model solving. A load balancing mechanism was introduced in the initialization phase to distribute the processes to AGVs and processing equipment as evenly as possible. The crossover probability was adaptively adjusted, and the algorithm was improved based on the processes, AGVs, and processing equipment.

[0130] Accordingly, this embodiment also provides an integrated scheduling system for multi-AGV green workshops based on an improved genetic algorithm, including:

[0131] The building module is used to construct a scheduling problem model for a multi-AGV green workshop. The model aims to minimize total energy consumption, maximum completion time and total cost, and includes energy consumption constraints, charging constraints and energy consumption constraints of processing equipment for AGVs in waiting, empty and full-load states.

[0132] The solution module is used to solve the scheduling problem model using an improved genetic algorithm; the chromosome encoding of the improved genetic algorithm includes process sequence encoding, equipment allocation encoding, and AGV allocation encoding.

[0133] The output module is used to output the optimal scheduling scheme through an improved genetic algorithm, and to schedule the AGVs and processing equipment in the workshop according to the output optimal scheduling scheme.

[0134] Example 2

[0135] The difference between the multi-AGV green workshop integrated scheduling method based on the improved genetic algorithm provided in this embodiment and the first embodiment is as follows:

[0136] To verify the effectiveness of the proposed algorithm, this embodiment was programmed using Matlab R2019b software. The experiment was conducted on a platform with an Intel(R) Core(TM) i5-12500 3.00 GHz processor and 16.0 GB of memory. The algorithm settings were Gen=1, GenMax=200, Nind=100, and P... c max =0.9, P c min =0.9, k = 2, P m =0.01, run 20 times and take the average value as a reference.

[0137] The effectiveness of the algorithm was verified using Brandimarte's MK01 and MK02 as benchmark examples. Both examples involved 6 devices, with 4 AGVs used for transport. The transport schedule of the AGVs between the material area, devices, finished products, and charging stations is shown in Table 1. In the table, St represents the material area, En represents the finished product area, Ch represents the charging station, and M1-M6 represent the device numbers. Relevant data for the AGVs and devices are shown in Table 2. Parameter C was set based on the operational conditions of a certain workshop. AGV =10, C shop =50.

[0138] Table 1. AGV transport schedule for MK01 and MK02 examples

[0139] Table 2 Relevant data for AGVs and equipment AGV no-load energy consumption AGV load energy consumption AGV waiting energy consumption AGV operating speed AGV maximum power AGV charging rate device standby power consumption Equipment processing energy consumption 1.2 3.6 0.1 1 100 20 1 1.5

[0140] The results verified using examples MK01 and MK02 are shown in Table 3, where Var represents the variance of the optimized results after 20 runs. Table 3 shows that the objective value optimized by IGA is better than that optimized by GA, and IGA improves optimization efficiency by 11.92% compared to GA. To further observe the convergence of the algorithm, Figure 5 and Figure 6 The iterative graphs for MK01 and MK02 in the examples are IGA and GA, respectively. Figure 5-6 As can be seen, IGA has better convergence performance and convergence speed.

[0141] Table 3 Optimization results of MK01 and MK02 examples under different algorithms

[0142] The configuration of the number of AGVs is a key decision variable affecting workshop scheduling performance and cost. Too few AGVs can easily become a bottleneck in workshop operations, leading to prolonged equipment waiting and completion times; too many AGVs increase fixed investment and maintenance costs. Therefore, it is necessary to analyze the optimal configuration number of AGVs in the workshop. Based on the analysis of the MK01 example, the optimization results with energy consumption as the objective are shown in Table 4. The table also shows the percentage of AGVs in different working states. Figure 7 As shown, from Figure 7 As can be clearly seen, the overall work efficiency (idle + full load) peaks with 2 AGVs, while the waiting rate reaches its lowest point with 2 AGVs (because with 1 AGV, the AGV will be working continuously, so the waiting time is the shortest). Afterward, as the number of AGVs increases, the waiting rate continues to climb, indicating a significant surplus of resources. Therefore, from a utilization perspective alone, configuring 2 AGVs is the most reasonable approach.

[0143] Table 4. Optimization results with energy consumption as the objective for different AGV configurations.

[0144] from Figure 8The data shows that configuring 5 AGVs is the optimal choice under the current conditions, achieving excellent cost-efficiency while keeping energy consumption within acceptable limits. Configuring 8 AGVs, on the other hand, involves sacrificing cost for maximum efficiency, and the more AGVs there are, the more difficult workshop management becomes.

[0145] The results of the example show that this method can provide workshop managers with different target options, enabling them to weigh key indicators such as energy consumption, maximum completion time and cost, as well as the configuration of AGVs based on specific production goals, thereby achieving comprehensive optimization of the workshop.

[0146] Note that the above description is merely a preferred embodiment of the present invention and the technical principles employed. Those skilled in the art will understand that the present invention is not limited to the specific embodiments described herein, and various obvious changes, readjustments, and substitutions can be made without departing from the scope of protection of the present invention. Therefore, although the present invention has been described in detail through the above embodiments, the present invention is not limited to the above embodiments, and may include many other equivalent embodiments without departing from the concept of the present invention, the scope of which is determined by the scope of the appended claims.

Claims

1. A multi-AGV green workshop integrated scheduling method based on an improved genetic algorithm, characterized in that, include: S1. Construct a scheduling model for a green workshop with multiple AGVs. The model aims to minimize total energy consumption, maximum completion time, and total cost. It also includes energy consumption constraints, charging constraints, and energy consumption constraints of processing equipment for AGVs in waiting, empty, and fully loaded states. S2. An improved genetic algorithm is used to solve the scheduling problem model; the chromosome encoding of the improved genetic algorithm includes process sequence encoding, equipment allocation encoding, and AGV allocation encoding. S3. Improve the genetic algorithm to output the optimal scheduling scheme, and schedule the AGVs and processing equipment in the workshop according to the output optimal scheduling scheme.

2. The multi-AGV green workshop integrated scheduling method based on an improved genetic algorithm according to claim 1, characterized in that, The objective functions for total energy consumption, maximum completion time, and total cost in step S1 are respectively expressed as follows: Where f1 represents the objective function of total energy consumption; f2 represents the objective function of maximum completion time; f3 represents the objective function of total cost; E operation E represents the energy consumption during equipment processing; idle E represents the energy consumption of the equipment when it is not under load. agv-wait E represents the energy consumption of the AGV while it is waiting; agv-idle E represents the energy consumption of the AGV when it is unloaded; agv-load Indicates the energy consumption of the AGV when it is fully loaded; C i Indicates the completion time of the workpiece; R represents the number of AGVs; r represents the current r-th AGV; C IJr Indicates the time when the AGV stops transporting the workpiece; C AGV Indicates; C shop This represents the cost incurred by the workshop per unit of time.

3. The multi-AGV green workshop integrated scheduling method based on an improved genetic algorithm according to claim 2, characterized in that, The procedure preceding step S2 also includes: The scheduling problem of multi-AGV green workshop is transformed into a single-objective optimization problem, and a comprehensive objective function is constructed.

4. The multi-AGV green workshop integrated scheduling method based on improved genetic algorithm according to claim 2, characterized in that, The charging constraints of the AGV include: when the AGV's power is below a threshold, if it is in a waiting or empty state, it will go directly to the charging station to charge; if it is in a fully loaded state, it will complete the current handling task first and then go to the charging station to charge.

5. The multi-AGV green workshop integrated scheduling method based on an improved genetic algorithm according to claim 1, characterized in that, In step S2, the improved genetic algorithm uses a load balancing mechanism to allocate processing equipment and AGVs to the process when initializing the population, so as to minimize the cumulative task load variance of all processing equipment and all AGVs.

6. The multi-AGV green workshop integrated scheduling method based on an improved genetic algorithm according to claim 5, characterized in that, In step S2, the selection operation of the improved genetic algorithm adopts a combination of roulette wheel selection and elite retention strategy, directly retaining a predetermined proportion of individuals with the best fitness in the population to the next population.

7. The multi-AGV green workshop integrated scheduling method based on an improved genetic algorithm according to claim 1, characterized in that, In step S2, the crossover operation of the improved genetic algorithm adopts an adaptive crossover probability. The adaptive crossover probability uses a non-linear decreasing strategy according to the iteration process, expressed as: Among them, P c P represents the crossover probability; cmin P represents the minimum crossover probability; cmax Gen represents the maximum crossover probability; Gen represents the current iteration number; GenMax represents the maximum iteration number; k represents the decay coefficient.

8. The multi-AGV green workshop integrated scheduling method based on an improved genetic algorithm according to claim 7, characterized in that, The crossover operation specifically involves: randomly selecting a target workpiece from the parent chromosomes, and for at least two discontinuous processes of the selected target workpiece, exchanging the equipment allocation gene segment and AGV allocation gene segment from the two parent chromosomes corresponding to the current process.

9. The multi-AGV green workshop integrated scheduling method based on an improved genetic algorithm according to claim 1, characterized in that, The mutation operation of the improved genetic algorithm in step S2 includes: randomly selecting a workpiece in the chromosome and exchanging the equipment allocation codes corresponding to any two processes within the selected workpiece; randomly selecting two gene positions in the AGV coding segment of the chromosome and exchanging the AGV numbers.

10. A scheduling system based on the multi-AGV green workshop integrated scheduling method based on the improved genetic algorithm as described in any one of claims 1-9, characterized in that, include: The building module is used to construct a scheduling problem model for a multi-AGV green workshop. The model aims to minimize total energy consumption, maximum completion time and total cost, and includes energy consumption constraints, charging constraints and energy consumption constraints of processing equipment for AGVs in waiting, empty and full-load states. The solution module is used to solve the scheduling problem model using an improved genetic algorithm; the chromosome encoding of the improved genetic algorithm includes process sequence encoding, equipment allocation encoding, and AGV allocation encoding. The output module is used to output the optimal scheduling scheme through an improved genetic algorithm, and to schedule the AGVs and processing equipment in the workshop according to the output optimal scheduling scheme.

Images