A method and system for optimizing production scheduling of multi-specification threaded steel bars parallel spraying

Abstract

The application discloses a method and system for optimizing production scheduling of threaded steel bars of multiple specifications in parallel spraying, the method comprising inputting an order queue containing threaded specifications, and pre-generating a standardized process file for each threaded specification; selecting N steel bars to form a parallel spraying batch, taking comprehensive cost as a target, introducing quality constraints, phase constraints and process parameter interlocking constraints, constructing a multi-constraint optimization model, and solving the optimal scheme by using a genetic algorithm to obtain an optimal scheme for production execution, and optimizing the model in combination with prediction and measurement results in the production process. The application provides a flexible mixed-flow production scheduling scheme of threaded steel bars of multiple specifications in balance of "cost-quality-efficiency" by constructing an intelligent production scheduling framework integrating order combination optimization, geometric layout optimization, process parameter optimization and quality prediction guarantee, and solves the production scheduling problem caused by workpiece geometric coupling, and provides key technical support for intelligent and flexible production of customized threaded steel bars of multiple specifications and small batches.

Description

Technical Field

[0001] This invention belongs to the field of intelligent manufacturing, automated production scheduling and process control technology, and relates to a production scheduling optimization method and system for parallel spraying of multiple specifications of threaded steel bars. Specifically, it relates to a parallel spraying production scheduling and process parameter collaborative optimization method and system applied to thermal spraying production lines, especially for mixed continuous production scenarios of multiple specifications (different diameters and thread parameters) of threaded steel bars. Background Technology

[0002] In industrial manufacturing, especially in metal surface treatment such as the production of anti-corrosion coatings for rebar by arc spraying, production scheduling and optimization are crucial for improving efficiency, reducing costs, and ensuring quality. Existing technologies have limitations in the following areas: (1) Traditional batching and scheduling methods for homogeneous batches: In many continuous or batch processing industrial scenarios, the core of scheduling is to batch orders with the same or similar attributes to optimize equipment utilization, reduce changeover costs, and balance logistics. Although these methods consider equipment capacity, process constraints, and logistics connections, their core is batching for homogeneous or substitutable products. The decision variables are mainly the composition, ordering, and allocation on parallel equipment. For parallel operation scenarios such as rebar spraying, which have strong geometric coupling and process interaction, traditional methods only consider the constraints of physical space and order delivery, and cannot handle the complex process quality problems introduced by geometric and physical factors such as the thread phase and edge effect of parallel rebars.

[0003] (2) Integrated Optimization Methods for Scheduling and Layout of Automated Production Lines: In the field of discrete manufacturing, such as CN114022028A, there are methods for the coordinated optimization of scheduling and equipment layout of hybrid production lines. These methods aim to minimize completion time and energy consumption by optimizing workpiece sequencing, the number and type of machine tools, and equipment spacing. Their constraints mainly include processing sequence, equipment exclusivity, and material handling time. However, these methods are suitable for scenarios where workpieces flow sequentially between equipment and processes are relatively independent. For special cases where multiple workpieces are processed simultaneously at the same station (spraying area) (parallel spraying), and the geometric relative positions between workpieces directly affect the processing effect (coating quality), existing scheduling and layout methods lack the ability to model and optimize the dynamic process coupling relationship between parallel processing units.

[0004] (3) Multi-specification material scheduling optimization method: In some material processing fields, such as the material cutting field described in CN117407966A, there are methods for optimizing the cutting of raw materials and required parts of multiple specifications in order to maximize material utilization. These methods use intelligent algorithms (such as genetic algorithms) to find the optimal material combination and cutting scheme. The core of these methods is to solve the one-dimensional or multi-dimensional knapsack problem and focus on the static combination optimization of material consumption. Although it involves "multi-specification", its decision and optimization objectives (length matching, minimum material waste) are not related to the real-time process interaction and quality risks caused by parallel operations in the dynamic production process, and cannot solve the quality consistency problem in parallel spraying.

[0005] Specifically, in the production of thermal spraying for threaded steel bars, current practices generally adopt two modes: (1) “Parallel production of batches of the same specification” mode (i.e., batching and scheduling method of homogeneous batches): steel bars of the same specification are grouped together for parallel spraying. This method requires stopping the machine to adjust the fixtures and process parameters when switching specifications, resulting in discontinuous production, low equipment utilization, and inability to flexibly respond to the production needs of multi-specification small batch orders.

[0006] (2) "Fixed fixture mixed specification production" mode: Attempts were made to spray steel bars of different specifications in parallel on fixed fixtures. However, due to the different thread geometry (pitch, rib height) and target coating thickness of steel bars of different specifications, a uniform spraying path and process parameters could not guarantee that the coating thickness and bonding strength of all steel bars, especially those at the edge, would meet the standards. At the same time, if the thread phase (rotation angle) of the parallel-placed threaded steel bars was not properly matched, "peak-to-peak" interference would occur, resulting in uneven distribution of the spray beam and seriously affecting the uniformity of the coating.

[0007] In summary, existing production scheduling optimization technologies have not addressed the unique challenges of parallel spraying production, which involves "multiple specifications, strong geometric coupling, and quality sensitivity." The industry lacks an intelligent production scheduling method that can dynamically and automatically make decisions within the order pool regarding: 1) the optimal parallel spraying rebar combination (specification and number of bars); 2) the optimal clamping position (clamp spacing) and initial circumferential rotation angle (thread phase) for each rebar; and 3) personalized matching and dynamic adjustment of spraying process parameters (especially edge compensation) for each rebar within the combination, in order to minimize overall production costs while meeting all hard constraints on rebar coating quality. Summary of the Invention

[0008] This invention proposes a scheduling optimization method and system for parallel spraying of multiple specifications of threaded steel bars, aiming to solve the scheduling problems caused by the mixed parallel production of multiple specifications of orders in the existing thermal spraying production of threaded steel bars. Specifically, these problems include: 1) the inability to achieve flexible mixed-flow production and efficient scheduling of multiple specifications of orders while ensuring that the coating quality (thickness, adhesion) of all parallel sprayed steel bars (especially steel bars of different specifications and positions) meets the standards; 2) the scheduling decision is disconnected from the setting of spraying process parameters, and the quality risks caused by geometric coupling (thread phase interference) and physical effects (spraying beam edge effect) during parallel operation are not considered; 3) existing methods are mostly experience-driven or single-objective optimization, which cannot achieve multi-objective collaborative optimization and closed-loop control of cost, efficiency and quality.

[0009] This invention is achieved through the following technical solution: This invention proposes a production scheduling optimization method for parallel spraying of multiple specifications of threaded steel bars, comprising the following steps: S1, Order Input and Process Preprocessing: Obtain an order queue containing different thread specifications, lengths L, target thicknesses T_target, and quantities Q, where thread specifications include diameter D, pitch P, and rib height H; pre-generate a corresponding standardized process file ProcessStd for each thread specification; S2, Construct a multi-constraint optimization model: Select N steel bars from the order queue to form a parallel spraying batch Bundle; minimize the overall cost C per unit length. TTL To achieve the objective, a multi-constraint optimization model is constructed. The decision variables of this optimization model include at least: the number of parallel reinforcement bars N, the selected reinforcement bar specification index combination X, the clamp spacing vector G between adjacent reinforcement bars, the initial circumferential rotation angle vector Φ for each reinforcement bar, and a flag vector Flag indicating whether edge compensation technology is enabled for each reinforcement bar. Edg The constraints of the optimization model include physical space constraints based on the effective travel of the worktable, hard quality constraints based on the online prediction model of coating quality, anti-interference constraints of thread phase, and interlocking constraints of process parameters. S3, Model Solving and Scheme Generation: The model is solved using an optimization algorithm to output the optimal parallel spraying scheme. The scheme specifically includes: the rebar arrangement layout, the initial circumferential rotation angle Φ of each rebar, and a personalized final process file (ProcessSet) matched for each rebar. The arrangement layout is determined by the specification index combination X and the fixture spacing vector G. The personalized final process file (ProcessSet) is based on a pre-generated standardized process file (ProcessStd) combined with a Flag. Edg The result is obtained after edge compensation correction; S4, Scheme Execution and Feedback Learning: Production clamping and spraying operations are carried out according to the optimal scheme; after production is completed, the online quality prediction data and offline measured data are compared, and the edge effect influence factor and / or phase interference influence factor used for quality prediction in the optimization model are dynamically adjusted according to the deviation, forming a self-learning closed loop of the optimization model.

[0010] Based on the above scheme, by integrating process parameter matching and quality prediction based on thread specifications into production scheduling decisions, flexible mixed-flow scheduling of multi-specification threaded steel bar orders can be achieved. While improving equipment utilization and production efficiency, this fundamentally ensures the coating quality of all steel bars in parallel spraying. Furthermore, a self-learning and adaptive mechanism for the production scheduling system is established, enabling the model to continuously accumulate production data and self-correct, adapt to process drift, and maintain the high quality and reliability of the production scheduling scheme in the long term.

[0011] Preferably, in step S1, the standardized process document ProcessStd includes dynamic spraying path parameters and a set of basic process parameters for a single rebar of a specific specification. The basic process parameter set includes conventional spraying control parameters such as spraying current, voltage, wire feed speed, compressed air pressure, and basic spraying distance. The dynamic spraying path parameters include the rebar rotation speed after compensating for thread shadows caused by the thread specification, the relative axial movement speed of the spray gun, and the dynamic spraying path. Based on this setting, standardized process benchmarks are provided for rebars of different specifications, ensuring that production scheduling optimization is based on high-quality individual processes through combination and adjustment, providing a fundamental guarantee for overall quality.

[0012] Preferably, in step S2, the comprehensive cost C TTL This includes material costs, energy costs, and quality risk penalty costs calculated based on an online coating quality prediction model. The quality risk penalty cost is positively correlated with the probability of the steel bar with the worst predicted bonding strength in the batch failing to meet quality standards. This quantifies quality risk into economic costs and incorporates them into the optimization objective, driving the algorithm to proactively seek high-quality, low-risk production scheduling solutions and achieving multi-objective collaborative optimization of quality and cost.

[0013] Preferably, in step S2, the online coating quality prediction model is an evaluation model that predicts the coating thickness and adhesion grade based on the unique temperature difference curves of the rib tops and valleys of the threaded rebar and their characteristic vectors for individual rebars with different thread specifications and locations; the online coating quality prediction model is used to obtain the probability that the coating adhesion of each rebar in the parallel spraying batch bundle is of a low grade. and with all the reinforcing bars As a hard constraint standard for quality, θ is the upper limit threshold of the probability of a "low" grade to ensure that the overall quality of the composite steel bars meets the standard. This eliminates batch scrapping and cost increases caused by insufficient bonding force during parallel spraying, shifting the optimization goal from simple efficiency to a win-win situation of "efficiency + quality". It prevents substandard combinations from being selected, fundamentally ensuring product quality consistency and achieving deep coupling of process, quality and production scheduling.

[0014] Preferably, in step S2, the thread phase anti-interference constraint means that for any two adjacent reinforcing bars in a batch, their initial circumferential rotation angle must satisfy... ; ; Where Pm and Pn are the pitches of the two reinforcing bars, Φm and Φn are their initial circumferential rotation angles, and δ is the preset phase difference threshold.

[0015] Furthermore, the phase difference threshold δ is set to a value range of [0.3, 0.7].

[0016] Based on the above design, the mathematical model forces the threads of the parallel reinforcing bars to be staggered, effectively avoiding spray beam reflection, material splashing and uneven coating thickness caused by "peak-to-peak" alignment, and solving the unique geometric interference problem of parallel spraying of threaded reinforcing bars.

[0017] Preferably, in step S2, the process parameter interlock constraint means that when the number of parallel sprayed bars N is greater than 1, the system automatically enables the edge compensation mode for the process parameters of the outermost rebar, and adjusts the gain of at least one of its wire feeding speed and spray gun moving speed, so as to actively compensate for the potential decline in coating quality of the outer rebar caused by the edge effect of the spray beam through the adaptive correction of process parameters, and ensures the consistency of coating quality of all rebars in the batch.

[0018] More preferably, the edge compensation mode is based on the preset standardized process file ProcessStd corresponding to the rebar to be compensated, increasing the wire feeding speed by 5% to 10%, and / or reducing the relative movement speed between the spray gun and the rebar by 5% to 10%, to achieve gain adjustment.

[0019] Preferably, in step S3, an improved genetic algorithm is used as the optimization engine, and the solution process includes: S301 adopts a fixed-length chromosome station hybrid coding, and the coding content includes: the number of parallel roots, the index of the steel bar specifications loaded at each station, the fixture offset, the initial circumferential rotation angle, and the edge compensation mark. S302, Import the fitness function Fitness=1 / (C TTL +C vio ), where C vioPenalties for violating the constraints; S303 uses customized crossover and mutation operators to iteratively optimize and obtain the optimal parallel spraying scheme that balances cost, quality, and efficiency. The mutation operation includes Gaussian mutation of the initial circumferential rotation angle decision variable.

[0020] Based on the above solution process, the genetic algorithm using hybrid encoding can effectively handle complex combinatorial optimization problems involving integer, continuous, and Boolean variables. It can efficiently search for high-quality production scheduling schemes through customized crossover and mutation operators. Furthermore, by performing a local fine search on the rotation angle, it is beneficial to find a better phase misalignment combination within the constraint space of thread phase anti-interference, thereby further improving coating uniformity.

[0021] This invention also provides a production scheduling optimization system for parallel spraying of multiple specifications of threaded steel bars, comprising: The order management module is used to receive and manage order queues containing different specifications and parameters; The process library module stores standardized process documents corresponding to various thread specifications. The production scheduling optimization engine is configured to execute any of the aforementioned methods to generate the optimal parallel spraying scheme; the production scheduling optimization engine is communicatively connected to the coating quality online prediction system, acquires the temperature measurement data required by the quality online prediction model in real time, and calls the model to perform quality risk assessment in the optimization calculation. The process issuance and execution module is used to issue the personalized process parameter package in the above scheme to the spraying equipment for execution; The quality feedback module is used to collect online quality data and drive the production scheduling optimization engine to perform model parameter self-learning.

[0022] Based on the above system, the method is integrated into the software and hardware system to realize the full-process automation and intelligent control from order input, intelligent scheduling, process self-matching to quality feedback closed loop, and to achieve deep integration of the scheduling system and the online quality inspection system, so that the scheduling decision is based on real-time and objective quality prediction data, ensuring the scientific nature and accuracy of the decision.

[0023] Compared with the prior art, the present invention has the following beneficial effects: (1) Pioneering a deep collaborative optimization model integrating "production scheduling-process-quality": Unlike traditional production scheduling that only considers the scheduling of orders, equipment, and time, or layout scheduling that only considers logistics and equipment utilization, this invention creatively incorporates personalized matching of spraying process parameters and online quality prediction results as core constraints and objectives into production scheduling optimization; through decision variables Φ and Flag EdgBy directly transforming process control methods such as thread phase adjustment and edge compensation into part of production scheduling decisions, it achieves integrated optimization from "what to schedule" to "how to schedule and how to spray," fundamentally solving the problem of quality consistency in parallel spraying of mixed specifications.

[0024] (2) A “phase anti-interference constraint” standard for the unique geometric coupling of threaded steel bars is proposed: By quantifying the serious impact of the geometric phase alignment of threaded steel bars on the coating quality during parallel spraying, it is creatively abstracted into a strict mathematical constraint condition: |sin(Δφ)|>δ, and it is incorporated into the production scheduling optimization as a key technical feature unique to parallel spraying of threaded steel bars. This is a completely new dimension that has never been involved in any existing general production scheduling method. By optimizing the initial circumferential rotation angle Φ, the quality defects caused by the “peak-to-peak” alignment of the threads are avoided in advance during the clamping stage, realizing the quality control concept of “prevention first” and reducing quality and material costs from the source.

[0025] (3) Constructing a dynamic risk penalty mechanism based on online quality prediction: Unlike post-event sampling or fixed rules, this invention introduces the online prediction probability of coating adhesion into the optimization objective function as the "quality risk penalty cost", which enables the optimization algorithm to actively avoid steel bar combinations and process schemes with poor prediction quality, and shifts quality control from "passive detection" to "active design", ensuring the quality reliability of the production scheduling scheme at the theoretical level.

[0026] (4) Realize a data-driven adaptive closed-loop production scheduling system: This invention uses production feedback to self-learn and update model parameters (such as edge effect factor and phase interference factor), enabling the system to adapt to changes in raw materials, environment, etc., and continuously iterate and optimize to form an intelligent production scheduling closed loop with continuous evolution capability.

[0027] In summary, the core technical effect of this invention is as follows: For the special and complex industrial scenario of parallel spraying of multiple specifications of threaded steel bars, it has for the first time constructed an intelligent production scheduling framework that integrates order combination optimization, geometric layout optimization, process parameter optimization, and quality prediction and assurance. This framework breaks through the boundaries of traditional production scheduling and scheduling technologies, solves the production scheduling problem caused by the direct coupling of workpiece geometric features with process results, and provides key technical support for the intelligent and flexible production of customized multi-specification, small-batch orders. Attached Figure Description

[0028] Other features, objects, and advantages of the present invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings: Figure 1 This is a schematic diagram of the workflow of the present invention. Detailed Implementation

[0029] The present invention will be further described in detail below with reference to embodiments, but the implementation of the present invention is not limited thereto.

[0030] Example 1 This invention provides a production scheduling optimization method for parallel spraying of multiple specifications of threaded steel bars, comprising the following steps: S1, Order Input and Process Pre-processing: S101, input all orders into the system and generate a dynamic order queue OrderList={O1,O2,…O i Each order O i Each contains: Thread specification (D) i ,P i H i ): The diameter D of the i-th reinforcing bar i Thread pitch P i With rib height H i ; L i : The length of the i-th rebar; Q i Order quantity; T_target i Target coating thickness; S102, Preprocessing: The system calls the preset path planning engine to process each thread specification (D) in the order queue. i ,P i H i A corresponding "Standard Optimized Process File" (ProcessStd) will be pre-generated. i ; ProcessStd is a standard process file pre-generated offline for a specific thread specification (uniquely determined by diameter D, pitch P, and rib height H). It contains all the basic process parameters required for spraying a single rebar of that specification (including conventional spraying control parameters such as spraying current, voltage, wire feeding speed, compressed air pressure, and basic spraying distance) and dynamic spraying path parameters. The dynamic spraying path parameters here can refer to the parameters generated by the relatively mature spraying dynamic path planning method in the prior art. Preferably, in this embodiment, the dynamic spraying path parameters include the rotation speed of the reinforcing bar, the axial movement speed of the spray gun relative to the reinforcing bar, and the spraying dynamic path that has taken into account the thread shadow compensation. The spraying dynamic path is generated as follows: based on the fact that the thread ribs and the spraying jet have an inclined angle during the rotation of the reinforcing bar relative to the spray gun, thus forming a thread shadow, the path planning is constructed by dynamically adjusting the rotation speed of the reinforcing bar and the axial conveying speed relative to the spray gun by calculating the thread shadow area in real time.

[0031] S2, Construct a multi-constraint optimization model: For a production batch to be optimized, select N (Nb) orders from the order queue. min ≤N≤N max The reinforcing bars are arranged in a parallel spraying batch bundle to minimize the overall cost per unit length. TTL To achieve the objective, a multi-constraint optimization model is constructed. The objective function, decision variables, and constraints of the multi-constraint optimization model are specifically designed as follows: (1) Optimize the objective function: minimize the overall cost per unit length. This includes material costs, energy costs, and quality risk penalty costs calculated based on an online coating quality prediction model. The quality risk penalty cost is positively correlated with the probability of the steel bar with the worst predicted bonding strength within the batch failing to meet quality standards. The specific calculation method is as follows:

[0032] in, Represents material costs, based on Calculate, where ρ is the density of the sprayed material. Let the coating volume of the i-th rebar be estimated based on its process parameters. Cost per unit volume of spray coating material; Represents energy consumption cost, according to calculate, , , These are the operating voltage, current, and spraying time for the spraying operation. Here, η represents the cost of electricity, and η is the energy conversion coefficient. Representing the cost of quality risk penalty: By invoking the online coating quality prediction model, the predicted probability of coating adhesion for the worst-quality reinforcing bars (typically edge bars) under this combination is evaluated. ,according to K is a very large penalty coefficient (e.g., 10). 5 This ensures that any substandard steel bar will render the plan unfeasible.

[0033] (2) Decision variables N: Number of parallel spray lines, N min ≤N≤N max The determination is made by comprehensively considering the number of parallel roots and the combination of specifications, as well as the minimum target overall cost and the introduction of quality constraints; X: An N-dimensional vector representing the thread specification (D, P, H) index of N steel bars selected from the order pool; G: Fixture spacing vector, an (N-1) dimensional vector, representing the adjustment amount of the center distance between any adjacent steel bars. As a continuous decision variable, it allows for optimization adjustment. As a decision variable, the basis is that in parallel spraying, the fixture spacing not only affects the space utilization rate but also the overlapping area of ​​the spray beam, which in turn affects the coating uniformity. Using it as an optimization variable allows for dynamic adjustment of the steel bar position, making the beam coverage more uniform while avoiding interference. This is one of the new dimensions that distinguishes it from traditional production scheduling. Φ: An N-dimensional vector representing the "initial circumferential rotation angle" of each rebar when it is loaded, which determines the orientation of the thread peaks and valleys during spraying and is used to optimize the thread phase; as the basis for decision variables: by optimizing Φ, the threads of parallel rebars can be staggered, eliminating the quality hazards caused by "peak-to-peak" or "peak-to-valley" alignment, introducing the geometric characteristics of threaded rebars into production scheduling optimization, and weakening the constraints of the inherent special characteristics of threaded rebar spraying on production scheduling optimization.

[0034] Flag Edg : The flag vector for edge compensation process, used to determine whether edge compensation process is enabled for the rebar at this station. It is represented by a Boolean value, taking the value 0 or 1. It serves as the basis for decision variables: When the rebar is at the outer station, due to the edge effect of the spray jet, its coating quality may be inferior to that of the middle rebar. By modifying the process parameters, the quality degradation caused by this edge effect can be compensated, and a balance can be achieved between quality and efficiency. If the middle rebar is already good enough, edge compensation can be omitted to save materials. If the edge quality prediction is poor, forced compensation is initiated to achieve "process self-adaptation" decision.

[0035] (3) Key constraints (a) Physical space constraints ; in, The effective travel in the width direction of the workbench is the maximum spraying width allowed by the equipment, which is also the limit distance from the center of the leftmost rebar to the center of the rightmost rebar. It is a fixed parameter determined by the hardware design. and These are the diameters of the reinforcing bars on both sides of the inner edge of the parallel spraying assembly (such as Φ16, Φ20, Φ25, etc.), which are known input parameters in the optimization model; The j-th clamp spacing is the center-to-center distance between two adjacent reinforcing bars, specifically the distance from the center of the j-th reinforcing bar to the center of the (j+1)-th reinforcing bar. It is one of the decision variables of the optimization model and is determined by the algorithm optimization. N represents the actual number of steel bars to be sprayed in parallel, the number of steel bars to be sprayed simultaneously in a batch; obtained by chromosome decoding.

[0036] (b) Hard constraints on quality:

[0037] in, To predict the probability value of the bonding strength of the i-th rebar being "low" by calling a pre-built online coating quality prediction model, θ is the upper limit threshold (e.g., 0.1) for the probability of the overall composite rebar meeting the "low" level, meaning that the predicted bonding strength of all rebars must be "low" below the threshold.

[0038] The reason for introducing hard quality constraints is that during parallel spraying, steel bars of different specifications and in different positions may have insufficient local bonding strength due to process differences (such as the coverage of the received paint, the position relative to the spray beam, etc.), which affects the overall quality compliance of all steel bars. Introducing quality constraints based on the probability of bonding strength at the "low" level can prevent substandard combinations from being selected, avoiding the "high output, low quality" problem that often occurs in the existing production scheduling method that only considers capacity. This shifts the optimization goal from simple efficiency to a win-win situation of "efficiency + quality", fundamentally ensuring the consistency of product quality.

[0039] (c) Thread phase anti-interference constraint For any two adjacent reinforcing bars within a batch, their initial circumferential rotation angles satisfy: ; ; in, , These are the thread pitches of the two reinforcing bars; , Its initial circumferential rotation angle; This indicates the rotation angle of the reinforcing bar m. Convert to the spatial phase of the rebar thread, pitch The axial period and rotation angle of the thread are defined. This determines the circumferential position of the current cross-section; It is a dimensionless value representing the relative offset of the two threads in the axial direction, expressed in radians; δ is a preset phase difference threshold, used to determine whether the thread phases are sufficiently misaligned. The preferred value range of the phase difference threshold δ is set to [0.3, 0.7]. More preferably, δ is 0.5, that is, the phase difference is 30°, 150°, 210°, 330°, etc. At this time, the peaks of the two threads of the adjacent steel bars are misaligned by more than 1 / 6 cycle, and the sprayed jet will not hit all the rib tops at the same time, nor will it be completely aligned with all the valleys.

[0040] The special and unique characteristics of introducing thread phase anti-interference constraints are as follows: (i) Solving the unique geometric coupling problem of threaded steel bars: The geometric feature of threaded steel bars is a complex curved surface composed of periodically raised ribs (thread peaks) and concave valleys (thread bottoms). When multiple such steel bars are placed side by side, the threads of adjacent steel bars form a "gear meshing" geometric relationship in space, which causes a coupling effect when they are sprayed in parallel. This constraint directly targets the special characteristics of threaded steel bars, quantifies the thread geometric features into mathematical constraints and incorporates them into the optimization model, so as to fully realize the refined control based on the special characteristics of geometric interference.

[0041] (ii) Preventing quality defects at the source and improving coating uniformity: The geometric coupling of threaded steel bars will cause the spray beam to hit all rib tops simultaneously when the threads of the two steel bars are completely in phase (peak to peak). This can result in excessive accumulation at the rib tops, increased obstruction at the valley bottoms, and material splashing onto the valley bottoms or outside of adjacent steel bars, leading to uneven coating thickness and material waste. Alternatively, if the threads of the two steel bars are completely out of phase (peak to valley), the spray beam may be guided by the rib tops to the valley bottoms, still resulting in asymmetrical beam distribution and affecting overall uniformity. By constraining the thread phase, the peaks and valleys of the threads of the parallel steel bars are staggered before spraying begins, promoting a more uniform distribution of the spray beam, improving the coating consistency of each steel bar, and preventing potential quality problems in advance, rather than eliminating them after inspection. This reflects the manufacturing philosophy of "prevention first".

[0042] (iii) Synergy with other constraints: Together with physical space constraints and quality constraints, it forms a complete constraint system to ensure that the optimization results are feasible in multiple dimensions such as geometry, quality and process.

[0043] (d) Interlock constraints of process parameters When the number of parallel sprayed bars N is greater than 1, the system automatically enables edge compensation mode for the process parameters of the 1-2 steel bars in the outermost station, and adjusts the gain of at least one of the wire feeding speed and spray gun moving speed.

[0044] The preferred settings for the edge compensation mode are as follows: based on the pre-defined standardized process file ProcessStd corresponding to the reinforcing bar to be compensated, adjust the wire feeding speed and the relative movement speed of the spray gun according to the actual position and diameter of the reinforcing bar. Slightly increase the overall wire feeding speed of the spray gun by 5-10% to compensate for material loss caused by edge effects; Slightly reduce the relative movement speed between the spray gun and the rebar by 5-10% to increase the deposition time in the edge area.

[0045] The special significance of introducing process parameter interlock constraints lies in the fact that the process adaptive strategy is used as part of the production scheduling decision, enabling the production scheduling scheme to achieve integrated optimization from "what to schedule" to "how to schedule and how to spray", thus proactively improving the edge and overall spraying quality.

[0046] Furthermore, the online prediction model for coating quality in the aforementioned process can employ existing prediction models capable of obtaining the bonding strength state; preferably, in this embodiment, the online prediction model for coating quality refers to: Using different thread specifications and the offset position relative to the spray beam center as variables, different groups of steel rebar samples were designed for spraying operations. Temperature data specific to the rib tops and valleys of the threaded steel rebar were collected and temperature difference curves were constructed. The feature vectors of these temperature difference curves were extracted as input, and a physical-guided deep learning model was used to predict the coating thickness and adhesion grade, including the probability that the coating adhesion of the steel rebar was "low". .

[0047] S3, Model Solving and Solution Generation: An improved genetic algorithm is used as the optimization engine to solve the model. The solution process includes: S301, Chromosome Coding: Under the condition of maximum parallel root count limitation, a fixed-length hybrid coding method is designed for each station, consisting of thread specification index X, fixture vector G, initial circumferential rotation angle Φ, and edge compensation flag Flag. Edg The chromosome encoding it constitutes; S302, Import fitness function: Fitness = 1 / (C TTL +C vio ), where C vio Penalties for violating the constraints; S303 employs customized crossover and mutation operators for iterative optimization to obtain the optimal parallel spraying scheme that balances cost, quality, and efficiency. Crossover refers to multi-point crossover in the "specification combination segment" and simulated annealing arithmetic crossover in the "process parameter segment." Mutation includes: low-probability increase / decrease mutations of N; Gaussian mutation of the initial circumferential rotation angle Φ to explore better thread phase combinations; and mutations of the fixture spacing vector G and Flag. Edg Perform adaptive mutation.

[0048] The output optimal parallel spraying scheme includes the rebar arrangement, the initial circumferential rotation angle Φ of each rebar, and the personalized final process file ProcessSet matched for each rebar. In the actual production process, the output files include: Parallel Spraying Operation Instruction: Includes steel bar arrangement layout (steel bar arrangement diagram, clamp spacing G), initial rotation angle Φ; Personalized Process Parameter Package: N different process files, based on the pre-generated standardized process file ProcessStd, with edge compensation instructions injected; Quality Prediction Report: Predicted thickness and bonding strength grade of each steel bar; Cost and Efficiency Report: Projected material consumption, energy consumption, and labor hours.

[0049] S4, Solution Implementation and Feedback Learning: During production, the robot or operator performs production clamping according to the above output documents, and the central control system sends the corresponding process parameter packages to the spraying system to execute the spraying operation. After production is completed, the predicted quality data obtained from the online coating quality prediction model is extracted, and the coating quality of the corresponding batch of steel bars is measured offline. The online quality prediction data after production is compared with the offline measured data. If a systematic deviation occurs (such as an inaccurate edge compensation coefficient), the "edge effect influence factor" and "phase interference influence factor" used for prediction in the optimization model are automatically updated to form a self-learning closed loop of the optimization model, making the next production scheduling optimization more accurate.

[0050] Among them, the "edge effect influence factor" refers to the degree of coating quality attenuation caused by the edge effect of the spraying beam on the outer station reinforcement involved in the aforementioned scheme, including parameters such as clamp vector, reinforcement diameter, and rib height; the "phase interference influence factor" refers to the coupling effect of the phase difference of adjacent reinforcement threads on coating quality, mainly referring to the initial circumferential rotation angle.

[0051] Example 2 This embodiment provides a production scheduling optimization system for parallel spraying of multiple specifications of threaded steel bars, including: The order management module is used to receive and manage order queues containing different specifications and parameters; The process library module stores standardized process documents corresponding to various thread specifications. The production scheduling optimization engine is configured to execute the method described in Embodiment 1 to generate the optimal parallel spraying scheme; the production scheduling optimization engine is communicatively connected to the coating quality online prediction system, obtains the temperature measurement data required by the quality online prediction model in real time, and calls the model to perform quality risk assessment in the optimization calculation. The process issuance and execution module is used to issue the personalized process parameter package in the above scheme to the spraying equipment for execution; The quality feedback module is used to collect online quality data and offline test data, and drive the production scheduling optimization engine to perform model parameter self-learning.

[0052] Combining Examples 1 and 2, a production scheduling optimization verification is performed using an actual production process implementation case, as detailed below: Order Pool: Import 100 steel bars of three specifications (Φ16, Φ20, and Φ25) as targets for spraying operations, and store the pre-generated standard optimized process files for each specification in the process library module.

[0053] Activate the production scheduling optimization engine and set the maximum number of parallel elements N based on the maximum travel in the width direction of the workbench. max =4, execute the multi-specification parallel spraying scheduling optimization method as in Example 1, and obtain N. max The chromosome codes for each workstation are: [(3,0,0,1),(5,10.5,157.3,1),(0,0,0,0),(5,-8.2,82.1,0)]. The meanings of each chromosome code are as follows: Workstation 1: Place a steel bar of specification 3, without offset or rotation, with edge compensation enabled; Station 2: Place a 5mm gauge rebar, offset it 10.5mm in the positive direction, rotate it 157.3°, and enable edge compensation; Workstation 3: Idle; Station 4: Place a size 5 rebar (another one), with a negative offset of 8.2mm and a rotation of 82.1°, without enabling edge compensation; The optimal production scheduling scheme output is as follows: Select Φ20 (2 pieces, specification 5) and Φ25 (1 piece, specification 3) from the order pool to form a batch, N=3; the optimal fixture spacing G=[28mm,30mm], the three steel bars have different initial rotation angles Φ=[0°,157°,82°], and the thread phases are staggered; enable edge compensation for the steel bar process of station 1 and station 2.

[0054] Verification results: Compared with the traditional parallel production method of the same batch, the production capacity of the present invention is increased by 25%, the material utilization rate is increased by 15%, and the coating quality of all specifications of steel bars meets the standards after sampling inspection.

[0055] As can be seen from the above, by implementing the method of the present invention, the deep integration of "process-quality-production scheduling" and the balance of materials, quality and efficiency are effectively realized. It breaks through the boundaries of traditional production scheduling and scheduling technologies, solves the production scheduling problem caused by the direct coupling of workpiece geometric features with process results, and provides key technical support for the intelligent and flexible production of customized multi-specification and small-batch orders.

[0056] The above description is merely a preferred embodiment of the present invention and does not constitute any limitation on the technical scope of the present invention. Therefore, any minor modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A production scheduling optimization method for parallel spraying of multiple specifications of threaded steel bars, characterized in that, Includes the following steps: S1, Order Input and Process Preprocessing: Obtain an order queue containing different thread specifications, lengths L, target thicknesses T_target, and quantities Q, where thread specifications include diameter D, pitch P, and rib height H; pre-generate a corresponding standardized process file ProcessStd for each thread specification; S2, Construct a multi-constraint optimization model: Select N steel bars from the order queue to form a parallel spraying batch Bundle; minimize the overall cost C per unit length. TTL To achieve the objective, a multi-constraint optimization model is constructed. The decision variables of this optimization model include at least: the number of parallel reinforcement bars N, the selected reinforcement bar specification index combination X, the clamp spacing vector G between adjacent reinforcement bars, the initial circumferential rotation angle vector Φ for each reinforcement bar, and a flag vector Flag indicating whether edge compensation technology is enabled for each reinforcement bar. Edg The constraints of the optimization model include physical space constraints based on the effective travel of the workbench, hard quality constraints based on the online coating quality prediction model, thread phase anti-interference constraints, and process parameter interlocking constraints. The thread phase anti-interference constraint means that for any two adjacent steel bars in a batch, their initial circumferential rotation angle must satisfy... ； ; in, , These are the thread pitches of the two reinforcing bars. , The initial circumferential rotation angle is δ, which is a preset phase difference threshold. The process parameter interlock constraint means that when the number of parallel sprayed bars N is greater than 1, the system automatically enables the edge compensation mode for the process parameters of the steel bar in the outermost station, and adjusts the gain of at least one of its wire feeding speed and spray gun moving speed. S3, Model Solving and Scheme Generation: The model is solved using an optimization algorithm to output the optimal parallel spraying scheme. The parallel spraying scheme includes: the arrangement of reinforcing bars, the initial circumferential rotation angle Φ of each reinforcing bar, and a personalized final process file ProcessSet matched for each reinforcing bar. The arrangement is determined by the specification index combination X and the fixture spacing vector G. The personalized final process file ProcessSet is generated by combining the pre-generated standardized process file ProcessStd with Flag. Edg The result is obtained after edge compensation correction; S4, Scheme Execution and Feedback Learning: Production clamping and spraying operations are carried out according to the optimal parallel spraying scheme; after production is completed, the online quality prediction data and offline measured data are compared, and the multi-constraint optimization model is calibrated and updated according to the deviation to form a self-learning closed loop.

2. The method according to claim 1, characterized in that, In step S1, the standardized process document ProcessStd contains dynamic spraying path parameters and a set of basic process parameters for a single steel bar of a certain specification. The dynamic spraying path parameters include the steel bar rotation speed, the relative axial movement speed of the spray gun, and the dynamic spraying path after compensating for the thread shadow generated by the thread specification.

3. The method according to claim 1, characterized in that, In step S2, the comprehensive cost C TTL This includes material costs, energy costs, and quality risk penalty costs calculated based on an online coating quality prediction model; the quality risk penalty costs are positively correlated with the probability of non-compliance of the steel bar with the worst predicted bonding strength within the batch.

4. The method according to claim 1, characterized in that, In step S2, the online coating quality prediction model is an evaluation model that predicts the coating thickness and adhesion grade based on the temperature difference curves of the rib tops and valleys of the threaded rebar and their eigenvectors for single rebars with different thread specifications and locations. This online coating quality prediction model is used to obtain the probability that the coating adhesion of each rebar in a parallel spraying batch is of a low grade. and with all the reinforcing bars As a hard constraint standard for quality, θ is the upper limit threshold of the probability of "low" grade to ensure that the overall quality of the composite reinforcement meets the standard.

5. The method according to claim 1, characterized in that, The phase difference threshold δ is set to a value range of [0.3, 0.7].

6. The method according to claim 1, characterized in that, The edge compensation mode is as follows: based on the preset standardized process file ProcessStd corresponding to the rebar to be compensated, the wire feeding speed is increased by 5% to 10%, and / or the relative movement speed between the spray gun and the rebar is reduced by 5% to 10%.

7. The method according to claim 1, characterized in that, In step S3, an improved genetic algorithm is used as the optimization engine, and the solution process includes: S301 adopts a fixed-length chromosome station hybrid coding, and the coding content includes: the number of parallel roots, the index of the steel bar specifications loaded at each station, the fixture offset, the initial circumferential rotation angle, and the edge compensation mark. S302, Import the fitness function Fitness=1 / (C TTL +C vio ), where C vio Penalties for violating the constraints; S303 uses customized crossover and mutation operators to iteratively optimize and obtain the optimal parallel spraying scheme that balances cost, quality, and efficiency. The mutation operation includes Gaussian mutation of the initial circumferential rotation angle decision variable.

8. A production scheduling optimization system for parallel spraying of multiple specifications of threaded steel bars, characterized in that, include: The order management module is used to receive and manage order queues containing different specifications and parameters; The process library module stores standardized process documents corresponding to various thread specifications. The production scheduling optimization engine is configured to execute the method as described in any one of claims 1-7 to generate the optimal parallel spraying scheme; the production scheduling optimization engine is communicatively connected to the coating quality online prediction system, acquires the temperature measurement data required by the quality online prediction model in real time, and calls the model to perform quality risk assessment in the optimization calculation. The process issuance and execution module is used to issue the personalized process parameter package in the above scheme to the spraying equipment for execution; The quality feedback module is used to collect online quality data and drive the production scheduling optimization engine to perform model parameter self-learning.

Images