A production and material distribution collaborative scheduling optimization method of intelligent assembly line

Abstract

The application provides a production and material distribution collaborative scheduling optimization method of an intelligent assembly line. The method is suitable for an assembly line scene: work-in-process is transported by a conveyor belt, and parts required by the work-in-process are transported to corresponding workstations by a feeding robot, and the feeding robot can access a charging station for charging. The method comprises the following steps: decomposing materials required for assembling each order into multiple material distribution tasks performed by the feeding robot according to information of the individualized order; searching for an optimal production sequence of the order and a sequence of the material distribution tasks performed by the feeding robot based on a Lagrangian relaxation technique and a subgradient framework; obtaining a charging instruction of the feeding robot according to the sequence of the material distribution tasks performed by the feeding robot; performing a neighborhood search on all the sequences of the material distribution tasks; and coordinating production activities of each workstation of the assembly line and material distribution and charging activities of the feeding robot according to the production sequence of the order and the execution sequence of the material distribution tasks.

Description

Technical Field

[0001] This invention relates to the field of intelligent manufacturing technology, specifically to a method for optimizing the collaborative scheduling of production and material distribution in an intelligent assembly line. Background Technology

[0002] Traditional assembly line production and material distribution scheduling schemes are formulated through a phased sequential decision-making process. Specifically, the production department schedules production (determining the production sequence of orders on the assembly line), and the logistics department formulates a material distribution plan (the delivery time of materials to each workstation on the assembly line) based on the production schedule. The production scheduling decision model used by the production department does not consider the spatial displacement process of materials; the material distribution decision model used by the logistics department treats the order production sequence as a predetermined parameter.

[0003] For traditional assembly lines that receive large volumes of orders and rely primarily on manual labor, this phased, sequential decision-making scheduling optimization method is highly effective. However, with increasingly personalized market demands, advancements in advanced technologies (such as robotics and the Internet of Things), and rising labor costs, intelligent assembly lines that combine multi-variety, small-batch production with high automation capabilities are replacing traditional assembly lines.

[0004] Traditional production and material distribution scheduling optimization methods for assembly lines cannot meet the operational management needs of intelligent assembly lines. The specific reasons are as follows:

[0005] (1) The intelligent assembly line uses feeding robots to perform material delivery tasks that were originally done by humans. The dimensions of material delivery decisions have increased. Considering the differences in three-dimensional size, quality, quantity of materials, as well as the load capacity and endurance of feeding robots, it is necessary to make more refined decisions for feeding robots than for workers (such as charging start and end times, the number of materials delivered in a single delivery, etc.).

[0006] (2) New material distribution constraints and decisions have led to increased differences in delivery time. The production sequence of different personalized orders is interdependent with the material delivery time. If the traditional phased sequential decision-making scheduling optimization method is used, the time saved in production cannot offset the time increased in logistics, and the total completion time will deviate from the optimal.

[0007] Based on the above reasons, it is necessary to propose a collaborative scheduling optimization method for production and material distribution suitable for intelligent assembly lines.

[0008] The disadvantages of existing technologies are as follows:

[0009] (1) The total system completion time obtained by making production and material distribution decisions in stages is suboptimal.

[0010] (2) The lack of adaptive charging and delivery volume decisions makes it impossible to guide the feeding robot to perform efficient charging and material delivery operations, which increases the idle rate of the feeding robot. Summary of the Invention

[0011] (a) Technical problems to be solved

[0012] To address the shortcomings of existing technologies, this invention provides a collaborative scheduling optimization method for production and material distribution in intelligent assembly lines. This method overcomes the shortcomings of traditional segmented sequential decision scheduling methods, which are prone to getting stuck in local optima and deviating from the global optimal solution. It considers the load capacity and battery capacity constraints of the feeding robot and outputs a collaborative scheduling scheme for production and material distribution that enables the total completion time of the system to reach the theoretical optimal level.

[0013] (II) Technical Solution

[0014] To achieve the above objectives, the present invention provides the following technical solution: a method for optimizing the collaborative scheduling of production and material distribution in an intelligent assembly line, specifically comprising the following steps:

[0015] S1. A orders were received, forming a set of orders. The demand vector for obtaining orders is D = {D1, D2, ..., D...} A The process calls the material distribution job generator and outputs a set of material distribution jobs. And Material Distribution Operation Index Table I;

[0016] S2. Initialize parameters: Lagrange multiplier vector λ = (λ1, λ2, ..., λ) F The current optimal average completion time Z best =+∞, the lower bound of the current mean completion time Z =-∞, step size α, threshold number of iterations for step size update Neighborhood search execution count The algorithm does not improve the upper bound of the number of iterations. And the counter for the number of iterations without algorithm improvement

[0017] S3. Based on vector λ, generate the execution order of V material delivery operations as follows:

[0018] A1. Solve the integer programming model for the following set covering problem using a column generation algorithm:

[0019] Objective function:

[0020] Constraints:

[0021] in Let r be the set of all material delivery operation sequences, with 0-1 parameter α. ir Let r represent whether material distribution operation i is performed, and let θ be a 0-1 decision variable. r Indicates whether r is selected as a solution to the model, c r Let r be the weighted total completion time of all material delivery operations, assuming i j For the j-th material delivery operation of r, n r Let r be the total number of material delivery operations.

[0022] A2. Solving this model yields V θ values. r For each material delivery operation sequence r with a value of 1, form a set of these material delivery operation sequences.

[0023] S4. Take them out one by one. The elements in the code are input into the logistics instruction generator to obtain the completion time of all material distribution operations.

[0024] S5. Based on vector λ, for each order a, calculate... Characterize the priority of order a, and then arrange the production order p of the orders from high to low priority;

[0025] S6. Call the production instruction generator, input p, output... and

[0026] S7. Calculate the subgradient vector g = t L -t P =(g1,g2,…,g F ),like but and Otherwise, let Reset

[0027] S8. Using the production sequence p as input, call... The next-neighbor searcher, after each call, yields... p and Input the cooperative instruction generator to obtain Z CF If Z CF <Z best Then Z best ←Z CE p best =p,

[0028] S9, if but Reset and use Update vector λ, that is

[0029] S10. Check the algorithm termination condition: If Then terminate the algorithm and return. Otherwise, proceed to step S4 to continue the iteration;

[0030] S11, will The execution sequence of the vth material delivery operation is r. v Assigning feeding robot v, and r v Input the logistics instruction generator to obtain the specific action instructions for the feeding robot v, and then input the S... best Input the collaborative instruction generator to obtain the production instructions for the assembly line.

[0031] Preferably, the weight of material delivery operation i in step A1 is λ. i .

[0032] Preferably, in step A1 Obtained by inputting r into the logistics instruction generator.

[0033] Preferably, in step S5, ρ a The larger the value, the higher the priority of order 'a'.

[0034] Preferably, in step S8, the neighborhood searcher is used to perform a neighborhood search of the material distribution operation sequence for a given production sequence, to obtain the neighborhood sequence of the material distribution operations. That is, the input is the production sequence p, and the output is a neighborhood set of V material distribution operation execution sequences.

[0035] Preferably, in steps S8 and S11, the collaborative instruction generator is used to coordinate the instructions output by the production instruction generator and the instructions output by the logistics instruction generator, thereby outputting the order start and completion times considering the impact of material delivery delays. That is, the input is a set consisting of the production sequence p and the execution order of V material delivery operations. Output: A vector of completion times for all orders, taking into account the impact of material delivery delays. The total completion time Z for all orders considering the impact of material delivery delays CF And the executable production instructions for the assembly line, including order start and finish times.

[0036] Preferably, in step S11, the logistics instruction generator is used to generate a logistics instruction that the feeding robot can execute based on the execution order of multiple material delivery operations. Input: nr The execution sequence of each material delivery operation, using Indicates that the output is n. r Completion time for each material delivery operation In addition, other logistics instructions that the feeding robot can execute include the time when the feeding robot arrives at the charging station, the start time of charging, the end time of charging, the time when it arrives at the picking point to pick up materials, which order's materials and corresponding quantities to pick up from which robotic arm workbench, which robotic arm workbench to arrive at and the arrival time and unloading completion time.

[0037] (III) Beneficial Effects

[0038] This invention provides a method for optimizing the collaborative scheduling of production and material distribution in an intelligent assembly line. Compared with existing technologies, it has the following advantages:

[0039] (1) The production and material distribution collaborative scheduling optimization method of this intelligent assembly line determines all logistics instructions of the feeding robot to be executed in the material distribution operation execution sequence in the logistics instruction generator, including instructions related to charging. As the undertaker of material distribution operations, the feeding robot must consider its load capacity and battery capacity constraints during the material distribution operation, and needs to combine the collaborative logic of production and material distribution in the intelligent assembly line scenario to issue relevant instructions. However, the classic material distribution method is based on the perspective of manual transportation or manual operation of vehicles, and cannot provide adaptive charging decisions for the feeding robot.

[0040] (2) The production and material distribution collaborative scheduling optimization method of this intelligent assembly line improves the material distribution sequence given the production sequence through a neighborhood searcher. This neighborhood searcher is based on the rule of "first-in, first-out" and implements the classic local search operation in a simpler way on this basis. Attached Figure Description

[0041] Figure 1 This is a logic diagram of the method of the present invention;

[0042] Figure 2 This is an abstract schematic diagram of the intelligent assembly system of the present invention. Detailed Implementation

[0043] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0044] Please see Figure 1-2 This invention provides a technical solution: a method for optimizing the collaborative scheduling of production and material distribution in an intelligent assembly line, specifically including the following steps:

[0045] S1. A orders were received, forming a set of orders. The demand vector for obtaining orders is D = {D1, D2, ..., D...} A The process calls the material distribution job generator and outputs a set of material distribution jobs. And Material Distribution Operation Index Table I;

[0046] S2. Initialize parameters: Lagrange multiplier vector λ = (λ1, λ2, ..., λ) F The current optimal average completion time Z best =+∞, the lower bound of the current mean completion time Z =-∞, step size α, threshold number of iterations for step size update Neighborhood search execution count The algorithm does not improve the upper bound of the number of iterations. And the counter for the number of iterations without algorithm improvement

[0047] S3. Based on vector λ, generate the execution order of V material delivery operations as follows:

[0048] A1. Solve the integer programming model for the following set covering problem using a column generation algorithm:

[0049] Objective function:

[0050] Constraints:

[0051] in Let r be the set of all material delivery operation sequences, with 0-1 parameter α. ir Let r represent whether material distribution operation i is performed, and let θ be a 0-1 decision variable. r Indicates whether r is selected as a solution to the model, c r Let λ be the weighted total completion time of all material distribution operations in r, and let λ be the weight of material distribution operation i. i Assuming i j For the j-th material delivery operation of r, n r Let r be the total number of material delivery operations. Obtained by inputting r into the logistics instruction generator;

[0052] A2. Solving this model yields V θ values. rFor each material delivery operation sequence r with a value of 1, form a set of these material delivery operation sequences.

[0053] S4. Take them out one by one. The elements in the code are input into the logistics instruction generator to obtain the completion time of all material distribution operations.

[0054] S5. Based on vector λ, for each order a, calculate... Characterize the priority of order a, and then arrange the production order p, ρ of the orders in descending order of priority. a The larger the value, the higher the priority of order 'a';

[0055] S6. Call the production instruction generator, input p, output... and

[0056] S7. Calculate the subgradient vector g = t L -t P =(g1,g2,…,g F ),like but and Otherwise Reset

[0057] S8. Using the production sequence p as input, call... The next-neighbor searcher, after each call, yields... p and Input the cooperative instruction generator to obtain Z CF If Z CF <Z best Then Z best ←Z CF p best =p,

[0058] S9, if but Reset and use Update vector λ, that is

[0059] S10. Check the algorithm termination condition: If Then terminate the algorithm and return. Otherwise, proceed to step S4 to continue the iteration;

[0060] S11, will The execution sequence of the vth material delivery operation is r. v Assigning feeding robot v, and r v Input the logistics instruction generator to obtain the specific action instructions for the feeding robot v, and then input the S... best Input the collaborative instruction generator to obtain the production instructions for the assembly line.

[0061] In this embodiment of the invention, in step S8, the neighborhood searcher is used to perform a neighborhood search of the material distribution operation sequence for a given production sequence, to obtain the neighborhood sequence of the material distribution operations. That is, the input is the production sequence p, and the output is a neighborhood set of V material distribution operation execution sequences. Workflow:

[0062] 1. Order Let V be a set of temporarily empty material delivery operation execution sequences, i.e., any None of these operations include any material delivery.

[0063] 2. Press p: = a1→a2→…→a A The order sequence will be as follows: The elements in are randomly assigned The elements in the array. For example: Suppose p := 1→2, First The elements in the array are randomly assigned to r1 and r2 in ascending order. For example, if 1 is assigned to r1, 2 to r1, and 3 to r2, then r1 := 2 and r2 := 1 → 3; then... The elements in the array are randomly assigned to r1 and r2 in ascending order. For example, if 4 is assigned to r2 and 5, 6, and 7 are equally assigned to r1, then r1: = 2→5→6→7 and r2: = 1→3→4.

[0064] In this embodiment of the invention, the collaborative instruction generator in steps S8 and S11 is used to coordinate the instructions output by the production instruction generator and the instructions output by the logistics instruction generator, thereby outputting the order start and completion times considering the impact of material delivery delays. The input is a set consisting of production sequence p and the execution order of V material delivery operations. Output: A vector of completion times for all orders, taking into account the impact of material delivery delays. The total completion time Z for all orders considering the impact of material delivery delays CF And the executable production instructions for the assembly line, including order start and finish times, and workflow:

[0065] 1. Remember Initialize the set consisting of the indices i of all material delivery operations corresponding to order a.

[0066] 2. Starting from the first order a1 in p, traverse in ascending order. If element i in the middle, but like (where i = I(a1,m,k)), then for all (where i' = I(a,m,k)), and then let... when Update after traversing all elements in the list. Then continue iterating. Execute the process until

[0067] 3. Update the vector obtained in step 2. And calculate

[0068] 4. According to t CF Taking into account the impact of material delivery delays, the start time of order a is...

[0069] In this embodiment of the invention, in step S11, the logistics instruction generator is used to generate a logistics instruction that the feeding robot can execute based on the execution order of multiple material delivery operations. Input: n r The execution sequence of each material delivery operation, using Indicates that the output is n. r Completion time for each material delivery operation And other logistics instructions that the feeding robot can execute, including the time when the feeding robot arrives at the charging station, the start time of charging, the end time of charging, the time when it arrives at the picking point to pick up materials, which order's materials and corresponding quantities to pick up from which robotic arm workbench, which robotic arm workbench to arrive at and the arrival time and unloading completion time, and the workflow:

[0070] 1. Instruct the feeding robot (r) to charge at the charging station at time 0, with a charging time of [time value missing]. At Arrive at the material collection point at the designated time to collect the material. After completing the material retrieval, the robot moves to the robotic arm workbench m1 corresponding to the material delivery operation i1 (i1 and m1 satisfy i1=I(a1,m1,k1)). Arrive at m1 at the appointed time Unload materials and complete material delivery operations at all times.

[0071] 2. If the material distribution operation i is known j The completion time is The subsequent material delivery instructions are: [to be continued] Arrive at the charging station to charge immediately; charging time is [time missing]. At Arrive at the material collection point at the designated time to collect the material. After completing the material pickup, the vehicle proceeds to the material delivery operation. j+1 The corresponding robotic arm worktable m j+1 (i j+1 and m j+1 Satisfy i j+1 =I(a j+1 ,m j+1 ,k j+1 )),At Time to arrive at m j+1 ,At Unload materials and complete material delivery operations at all times. j+1 .

[0072] 3. Based on step 1, the completion time of all material delivery operations on r can be obtained by following step 2.

[0073] like Figure 2 As shown, the system consists of an intelligent assembly line (containing M robotic arm workstations, each responsible for assembling one component of a product), a component pickup point, V feeding robots, and a charging station (containing V charging positions located near the component pickup point). The system can assemble components into the finished product required by the customer based on order information. Let a, m, and v represent the sets of orders, robotic arm workstations, and feeding robots, respectively. Let a, m, and v represent the indices of the orders, robotic arm workstations, and feeding robots, respectively. Due to the load capacity constraint of the feeding robots, the materials required for order a at location m on the robotic arm workstation may require multiple deliveries to meet the demand. Therefore, let v be an index of the order. Let k represent the set of delivery batches of order a at the robotic arm workbench m, and let k denote the index of the delivery batch in this set. The system parameters include:

[0074] 1. The cycle time T of the intelligent assembly line CT (s), conveyor belt speed γ c (m / s), mold change time T MC (s);

[0075] 2. The coordinates (x, y) of the m-th robotic arm worktable m ,y m );

[0076] 3. Coordinates of the charging station (x) c ,y c );

[0077] 4. Coordinates of the material collection point (x o ,y o );

[0078] 5. The load capacity C (kg), battery capacity B (kWh), travel speed γ (m / s), charging rate δ (kWh / s), power consumption rate Δ (kWh / m), and material handling time T of each feeding robot. LO (s), unloading time T UN (s);

[0079] 6. The time it takes for the conveyor belt to transfer the workpiece from the robotic arm workbench m to the robotic arm workbench m' The feeding robot starts from point j (coordinates (x) j ,y j Move to point j' (coordinates (x) j' ,y j' Travel time

[0080] This method embeds a material distribution operation generator, a production instruction generator, a logistics instruction generator, a collaborative instruction generator, and a neighborhood searcher. Their respective functions and detailed processes are as follows:

[0081] Material delivery operation generator

[0082] • Function: Break down the materials required for all orders into multiple material delivery operations.

[0083] Input: Information on all orders, including the order collection. Order Required product quantity D a ,Order The required product has a mass w of material at position m on the robotic arm's worktable. a,m .

[0084] Output:

[0085] 1. A collection of material distribution operation numbers A set of natural numbers starting from 1, where each natural number represents a material delivery operation.

[0086] 2. A material distribution operation index table I: Given the order a, the robotic arm workstation m, and the batch k to which the material belongs, the index table I of the material distribution operation can be obtained. This table contains the corresponding elements. The purpose of this table is to facilitate the retrieval of information about material distribution operations, as each operation corresponds to three key pieces of information: which order the delivered materials serve (corresponding to 'a'), which robotic arm workbench they are assembled on (corresponding to 'm'), and which batch (corresponding to 'k'). Index table I provides two ways to reference material distribution operations, one of which is using a set... The element in the text represents material distribution operation i, and the second is...<a,m,k> A triple consisting of three pieces of information is called a material distribution operation.<a,m,k> , when i=I(a,m,k),<a,m,k> The 'i' and 'i' represent the same material delivery operation.

[0087] 3. The vector q = (q1, q2, ..., q...) represents the material transportation volume corresponding to the material distribution operation. F ), where qi represents the amount of material that the material delivery operation i = I(a,m,k) should transport at the workbench m of the robotic arm for order a.

[0088] Workflow:

[0089] 1. According to w a,m Calculate the number of material delivery batches required for order a at the robotic arm workbench m. make Let k represent the set of material delivery batches required for order a at the robotic arm workbench m.

[0090] 2. Number the material delivery operations. There are a total of Each material distribution operation is numbered from 1 to F. This represents a set of material distribution operations.

[0091] 3. Generate index table I for material distribution operations. Inputting a, m, k, index table I returns the material distribution operations in the set... The corresponding element in the data is represented by i = I(a,m,k). The material distribution operation number.

[0092] It has the following characteristics:

[0093] i. If 'a' is different, then the larger 'a' is, the larger 'i' is;

[0094] ii. If a is the same and m is different, then the larger m is, the larger i is;

[0095] iii. If a and m are the same, then the larger k is, the larger i is.

[0096] 4. Calculate q i The detailed formula is: when 1≤k≤K a,m At -1, When k = K a,m hour, ).

[0097] Production instruction generator

[0098] • Function: Generates a series of production instructions that can be executed on the assembly line based on the assembly sequence of the order.

[0099] Input: The assembly sequence of A orders, expressed as p:=a1→a2→…→a A .

[0100] Output:

[0101] 1. A vector consisting of order completion times in This indicates the completion time of order a.

[0102] 2. Based on vectors The order start time is obtained by reverse calculation.

[0103] 3. A vector consisting of the latest completion times of material delivery operations without causing production delays. in This indicates the latest completion time of material delivery operation i without causing delays in production.

[0104] Workflow:

[0105] 1. Calculate the completion time of order a1.

[0106] 2. If order a is known j The completion time is Then order a j+1 The completion time is

[0107] 3. Based on step 1, step 2 can be used to obtain the completion time of all orders, and then vectorized. Output.

[0108] 4. Given We can deduce the start time of order a by working backwards.

[0109] calculate make Indicates all The vector formed by these vectors.

[0110] Based on the aforementioned parameters, three instruction generators, and a neighborhood searcher, and using order information as input, this method can output a production and material distribution collaborative scheduling scheme S with the objective of minimizing the total completion time of all orders. best S best The core is the production sequence p best and V sets of material delivery operation execution sequences Record for use Specific instructions related to production and material distribution coordination, such as production instructions for the assembly line (including the start and completion times of each order) and material distribution instructions for all feeding robots (including when the feeding robot visits the charging station, when it starts charging, when it ends charging, when it visits the picking point, when it picks up materials, which order's materials are picked up from which robotic arm workbench, how much material is picked up, when it visits which robotic arm workbench, and when it unloads, etc.), can be expressed in S. best The input is obtained by calling the collaborative instruction generator, production instruction generator, and logistics instruction generator. This method is based on the Lagrange relaxation technique and the subgradient algorithm framework. Figure 2 A logic diagram of the method is shown.

[0111] from Figure 1 As can be seen, this method employs the Lagrange relaxation technique (i.e., introducing the Lagrange multiplier vector λ) and the subgradient method framework (i.e., iterating over the Lagrange multiplier vector λ using the subgradient vector g). This method converges to a stable target value. Z ( Z For Z CF (a lower bound) and terminates. In this method, the Lagrange multiplier λ i It indicates The value of λ is the value of the materials delivered by material delivery operation i that arrive at their corresponding robotic arm workbench before the start of assembly. i This is considered the priority of material delivery operation i. λ i The larger the value, the earlier the material delivery operation i should be completed. Therefore, if based on λ... Greater than This material delivery operation i causes a delay, after which... The update, λ i The value will increase (because) It is a positive value. (If positive), the increased λ i This will make it possible in the next iteration Decrease, and at the same time, due to the priority of order a. It will decrease, making Increase, The reduction and The increase will maximize and Towards Trend correction. Conversely, if based on λ... Less than This material delivery operation i will not cause a delay, but its priority can be lowered to allow higher-priority material delivery tasks to be completed earlier. The update, at this time λ i The value will decrease (because) It is a positive value. (If the value is negative), the decreased λ i This will make it possible in the next iteration Increase, at the same time, due to the priority of order a It will increase, making Decrease The increase and The reduction will minimize and Towards The trend is corrected. After multiple iterations, the vector λ will converge to a stable value, so that for all material delivery tasks, due to... It has the least impact on the overall completion time of the final system.

[0112] In the output vector t L During the process, the logistics instruction generator developed in this method can output various instructions to control the operation of the feeding robot, especially the charging instructions. Moreover, these charging instructions can adaptively adjust according to the execution sequence of material delivery operations, with the aim of maintaining the feeding robot's power at a "just right" level (neither causing the material delivery time to be too late due to excessively long charging time, nor causing the feeding robot to be unable to complete the material delivery operation and return to the charging station due to insufficient power due to excessively short charging time).

[0113] Furthermore, the neighborhood searcher developed in this method can search for a given neighborhood. Local adjustments are made to search for a better solution given the production order, thus improving the quality of the final output solution.

[0114] In summary, the production and logistics distribution collaborative scheduling optimization algorithm provided by this invention integrates the originally independently managed production and material distribution processes, and takes into account the constraints of the loading capacity and endurance of the feeding robot that are ignored in traditional methods. It can simultaneously output collaborative production instructions and material distribution instructions, and obtain a better total order response time than traditional methods, providing automated production and material distribution collaborative decision support for the operation and management of intelligent assembly lines.

[0115] Furthermore, any content not described in detail in this specification is existing technology known to those skilled in the art.

[0116] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0117] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for optimizing the collaborative scheduling of production and material distribution in an intelligent assembly line, characterized in that: Specifically, the following steps are included: S1. A orders were received, forming a set of orders. Obtain the demand vector of orders. Call the material distribution job generator to output a set of material distribution jobs. and material distribution operation index table ; S2. Initialization parameters: Lagrange multiplier vector Current optimal average completion time The lower bound of the current average completion time Step length The threshold for the number of iterations for step size updates. , number of neighborhood search executions The algorithm does not improve the upper bound of the number of iterations. And the counter for the number of iterations without algorithm improvement ; S3, based on vector Generate as follows The execution sequence of each material delivery operation: A1. Solve the integer programming model for the following set covering problem using a column generation algorithm: Objective function: ; Constraints: , , ; in Sequence of all material delivery operations A set of 0-1 parameters express Is it a material delivery operation? 0-1 decision variables express Whether it is selected as a solution for the model. for The weighted total completion time of all material delivery operations, assuming , for The Material delivery operations for The total number of material delivery operations on the platform, then ; A2. Solving this model yields... indivual Material delivery sequence corresponding to a value of 1 This will form a set of material delivery operation sequences. ; S4. Take them out one by one. The elements in the code are input into the logistics instruction generator to obtain the completion time of all material distribution operations. ; S5. Based on vectors For each order ,calculate Characteristic Orders The priority is then determined, and the production sequence of orders is arranged from highest to lowest priority. ; S6. Call the production instruction generator and input... Output and ; S7. Calculate the subgradient vector ,like ,but and Otherwise, let Reset ; S8, in production sequence For input, call The next-neighbor searcher, after each call, yields... ,Will and Input the cooperative instruction generator to get ,like ,but , , , ; In step S8, the neighborhood searcher performs a neighborhood search on the material distribution operation sequence for a given production sequence to obtain the neighborhood order of the material distribution operations. The input is the production sequence. Output: The neighborhood set of material delivery operation sequences constituted by the execution sequence of each material delivery operation. ; The collaborative instruction generator is used to coordinate the instructions output by the production instruction generator and the logistics instruction generator, thereby outputting the order start and completion times that take into account the impact of material delivery delays. The input is the production sequence. , A set consisting of the execution sequence of individual material distribution operations. Output: A vector of completion times for all orders, taking into account the impact of material delivery delays. Total completion time for all orders, taking into account the impact of material delivery delays. And the executable production instructions for the assembly line, including order start and finish times; S9, if ,but Reset and use Update vector ,Right now ; S10. Check the algorithm termination condition: If If the condition is met, the algorithm terminates and returns. Otherwise, proceed to step S4 to continue the iteration; S11, will The Middle Execution sequence of material delivery operations Distributing and feeding robots ,Will Input the logistics instruction generator to obtain the feeding robot. Specific action instructions will Input the collaborative instruction generator to obtain the production instructions for the assembly line; In step S11, the logistics instruction generator is used to generate a logistics instruction that the feeding robot can execute based on the execution order of multiple material delivery operations. Input: The execution sequence of each material delivery operation, using Indicates, output: Completion time for each material delivery operation , In addition, other logistics instructions that the feeding robot can execute include the time when the feeding robot arrives at the charging station, the start time of charging, the end time of charging, the time when it arrives at the picking point to pick up materials, which order's materials and corresponding quantities to pick up from which robotic arm workbench, which robotic arm workbench to arrive at and the arrival time and unloading completion time.

2. The method for optimizing the collaborative scheduling of production and material distribution in an intelligent assembly line according to claim 1, characterized in that: The material delivery operation in step A1 The weight is .

3. The method for optimizing the collaborative scheduling of production and material distribution in an intelligent assembly line according to claim 1, characterized in that: In step A1 By Obtained by inputting logistics instructions into the generator.

4. The method for optimizing the collaborative scheduling of production and material distribution in an intelligent assembly line according to claim 1, characterized in that: In step S5 The larger the value, the more orders... The higher the priority, the better.

Images