Small Batch Production Mode Research Articles

Quality prediction for rocket body structure manufacturing process based on improved ELM

In the current multi-variety and small-batch production mode, this paper proposes a quality prediction method for rocket body structure manufacturing process based on improved extreme learning machine (ELM). Aiming at the non-optimal problems caused by random selection of input layer weights and hidden layer bias of ELM, particle swarm optimization (PSO) is combined with gravitational search algorithm (GSA) to form a hybrid algorithm to optimize the random parameters of ELM. Taking a weld in a rocket body structure as an example, the weld width prediction model based on improved ELM is established and verified. The results show that compared with the prediction model based on the standard ELM and the prediction model based on PSO-ELM, the prediction model based on improved ELM has higher prediction accuracy, and its prediction deviation does not exceed 2.5%, which can effectively predict the weld width.

Multi-variety and small-batch production quality forecasting by novel data-driven grey Weibull model

With the coming of intelligent manufacturing, multi-variety and small-batch production mode has gradually become popular. Aiming at the characteristics of high dimensional information and limited samples in this mode, a novel data-driven grey Weibull model is established for product quality prediction. Firstly, a high-dimensional information from the production process is integrated as the process variation using a data-driven method. Then, the quality prediction function is deducted by considering that process variation follows Weibull distribution-based mechanism, the Hausdorff difference scheme is adopted to weaken the error from the difference to the differential, heuristic algorithm is selected to optimize the distribution parameters of the model. Finally, an experimental analysis is designed using the dataset from some personalized customization manufacturer in China. Results show that the proposed model is not only superior to the other eight models in terms of stability and prediction accuracy, but also boasts the features of amalgamation of data-driven and mechanism-driven methods, which can simultaneously process high-dimensional information and limited samples in multi-variety and small-batch production system.

A matheuristic for flexible job shop scheduling problem with lot-streaming and machine reconfigurations

Multi-variety and small-batch production mode enables manufacturing industries to expeditiously satisfy customers' personalised demands, where a large amount of identical jobs can be split into several sublots, and be processed by reconfigurable machines with multiple machining technics. However, such highly flexible manufacturing environments bring some intractable problems to the production scheduling. Mathematical programming and meta-heuristic methods become less efficient when a scheduling problem contains both discrete and continuous optimisation attributes. Therefore, matheuristic, which combines advantages of the two methodologies, is regarded as a promising solution. This paper investigates a flexible job shop scheduling problem with lot-streaming and machine reconfigurations (FJSP-LSMR) for the total weighted tardiness minimisation. First, a monolithic mixed integer linear programming (MILP) model is established for the FJSP-LSMR. Afterwards, a matheuristic method with a variable neighbourhood search component (MH-VNS) is developed to address the problem. The MH-VNS adopts the classical genetic algorithm (GA) as the framework, and introduces two MILP-based lot-streaming optimisation strategies, LSO1 and LSO2, to improve lot-sizing plans with varying degrees. Four groups of instances are extended from the well-known Fdata benchmark to evaluate the performance of proposed MILP model, LSO1 and LSO2 components, and MH-VNS. Numerical experimental results suggest that LSO1 and LSO2 are efficient in different scenarios, and the proposed MH-VNS can well balance the solution quality and computational costs for reasonably integrating the GA- and MILP-based local search strategies. In addition, a complicated FJSP-LSMR case is abstracted from a real-world shop floor for processing large-sized structural parts to further validate the MH-VNS.

A multi-dimensional cognitive framework for cognitive manufacturing based on OAR model

With the production system shifting to a multi-variety and small-batch production mode, the production process faces more user requirements, changes, and uncertainties. To solve the above problems, it is necessary to obtain the status and trend changes information and provide information support for the optimization of decision-making and dynamic adjustment of the production system. However, the production system cognition faces the problems of state coupling, state dynamic transfer and transition, and multi-system interweaving, which makes the production system cognition face huge challenges. Combining technologies such as the Internet of Things, industrial big data, and artificial intelligence, cognitive manufacturing can realize dynamic cognition of the production process, support dynamic adjustment, and become a promising way to solve the dynamic changes and uncertainties of production systems. In addition, as a formal expression of information processing and knowledge learning process in cognitive informatics, the Object-Attribute-Relation (OAR) model can effectively guide the construction of the production process cognitive mechanism. Therefore, this paper proposes a multi-dimensional cognitive framework based on OAR model of the human cognitive world for the dynamic cognitive needs of production system. The framework carries out dynamic cognition from the three dimensions of the manufacturing unit, production situation, and production system, and builds the continuous cognitive abilities from the three dimensions of analysis, decision-making, and learning. By integrating intelligent algorithms in the fields of artificial intelligence, a computable digital twin model is constructed as a carrier to provide the cognitive enabling technologies and capabilities for the production system. Finally, the feasibility of the proposed framework is illustrated by the developed computational digital twin platform. The computable digital twin platform provides the production system with important cognitive capabilities such as states perception, trend prediction, optimization decision-making, and knowledge learning, to support the dynamic cognition and optimization decision-making of the production system, and lay a technical foundation for adaptive production and cognitive manufacturing.

Data-Driven Virtual Metrology and Retraining Systems for Color Filter Processes of TFT-LCD Manufacturing

With the innovations in commercial-electronics products and fierce competition in the global panel industry, most panel manufacturers have adopted the small batch production mode to deal with a wide range of customization. With small batch sizes, productive yield and troubleshooting are considered the top priority, so monitoring capability conducted via virtual metrology (VM) is essentially necessary to satisfy the requirement of process quality. In this paper, the real-world dataset involving the photoresist (PR) spacer heights of color filter (CF) in the array sector of TFT-LCD manufacturing is investigated. In practice, the PR spacer heights can only be measured in an infrequent manner due to the scheduling restriction. Without taking additional sample measurements, how to design a high-accuracy VM model based on small batch sizes warrants urgent research for the TFT-LCD industry. The proposed framework can be divided into two parts. First, two novel distance-measuring methods are proposed, Direct (fully-connected) Neural Network (DNN) and K-means, to allocate the most affiliated positions so that the VM system can utilize these positions to execute metrology prediction on the same product. Next, a modified random forests (RFs) regression model is integrated into the VM system to create an ensemble VM predictor that can handle multiple products of different small batch sizes. Lastly, the VM model is repeatedly evaluated, to check the performance of the re-training procedure when the model needs to be renewed between production maintenance periods.

Pattern Recognition of Quality Control Chart of Multi-variety and Small-batch Production Mode Based on MC-GA Optimized BP

To solve the problem of product quality control in multi-variety and small-batch production mode, a quality control chart recognition model of MC-GA optimized BP was established by using Monte Carlo method to simulate the product quality data characteristics of different varieties and combining the data processing capability of neural network and the adaptive global optimization search capability of genetic algorithm; then the MC(Monte Carlo method)-GA(Genetic Algorithm) optimized BP(Back Propagation Neuron Network) was compared with the traditional MC-BP is compared and analyzed in terms of accuracy, time and error in quality control graph pattern recognition, and it is concluded that the former is better in quality control graph pattern recognition. It can be seen that the MC-GA optimized BP model is more effective in identifying the quality status of products in the multi-variety and small-batch production mode, and can accurately predict the stability of the manufacturing process, so as to enhance the competitiveness of enterprises.

Research on quality warning of PZT sintering process based on SPC

In the small-batch production mode of multiple varieties of piezoelectric ceramics, according to the special requirements of batch sintering of the same batch of products in the sintering process, as well as the lack of quality sample data, quality status monitoring and inaccurate early warning, a combined process quality monitoring method of general control chart and CUSUM control chart was proposed. The method can effectively identify the slight deviation of quality and realize the early warning of process quality. The feasibility of this method is verified in this paper.

Multi-Objective Optimization of Workshop Scheduling with Multiprocess Route Considering Logistics Intensity

In order to solve the problems of flexible process route and workshop scheduling scheme changes frequently in the multi-variety small batch production mode, a multiprocess route scheduling optimization model with carbon emissions and cost as the multi-objective was established. At the same time, it is considered to optimize under the existing machine tool conditions in the workshop, then the theory of logistics intensity between equipment is introduced into the model. By designing efficient constraints to ensure reasonable processing logic, and then applying multilayer coding genetic algorithm to solve the case. The optimization results under single-target and multi-target conditions are contrasted and analyzed, so as to guide enterprises to choose a reasonable scheduling plan, improve the carbon efficiency of the production line, and save costs.

Semi-dynamic maintenance scheduling for multi-station series systems in multi-specification and small-batch production

This paper proposes a new semi-dynamic Opportunistic Maintenance (OM) policy for multi-station series systems, with considering the changing production environment brought by the multi-specification and small-batch production mode. It combines the advantages of the time-window-based OM policy, which can decrease the complexity of the solution space for PM grouping, and the dynamic OM policy, which is capable of adapting to the changing production environment. Whenever one of the stations reaches its original Preventive Maintenance (PM) moment, an OM decision cycle of the system begins. During this decision cycle, all the stations, whose original PM moments are within the presented dynamic Maintenance Time Window (MTW), form a PM group. This dynamic MTW is different for each station, and it is obtained based on the predetermined static MTW and the dynamic impacts from the current production task. The static MTW is constant, and it is derived according to the expected production plan without considering the changing production environment. To balance the maintenance cost for the obtained PM group, the PM is not performed right at the time when one of the stations reaches its original PM moment, and it is postponed. The optimal duration of the postponement is dynamically obtained by minimizing the total maintenance cost within this duration. Finally, numerical examples and comparisons are illustrated to show the effectiveness of the proposed semi-dynamic OM policy.

Modular Design of NC Program in Multi-varieties and Small-batch Production Mode

Aiming at design efficiency and quality problems of NC program in multi-varieties and small-batch production mode, this paper applied the concept of product modular design into NC program design, and presented the modular design method of NC program and its implementation techniques. By design the module NC program structure and build the database of basic module, provided technical support for the modular design of NC program. The practical application shows those: the constructed module database can be reused and can be combined, using the modular design method can effectively improve the design efficiency and quality of NC program, especially for multi-varieties and small-batch parts, provides an optimal way for NC program to quickly meet the individual demands.

An automatic scheduling method for weaving enterprises based on genetic algorithm

As the multi-varieties and small-batch production mode become more popular in weaving enterprises, the traditional manual operated scheduling exposes the disadvantages of low work efficiency and unsatisfying result. In this paper, by summarizing the weakness previous model, a more practical optimization model is developed for weaving production scheduling to reduce the schedulers’ labor. The model describes the optimization of warp beam looming schedule in weaving process based on the analysis of the schedulers’ working process in weaving enterprises. Genetic algorithm (GA) is adopted to solve the model. To improve the computational efficiency and avoid prematurity convergence of GA, the termination condition is updated and the genetic parameters are optimized based on statistical data. The improved GA with optimized parameters gets solutions superior to manual scheduling with a quicker convergence, which has great practical value.

Challenges Posed by Multi Variety and Small Batch Production Mode and Countermeasures Acted by Chinese Small and Medium Enterprises

Multi variety and small batch production mod has brought such challenges to China's small and medium enterprises as more difficult of production planning increased difficulty of material requirement planning, and difficult of clearance after order completion. How to adapt to the developments and changes in the market and respond effectively to "multi variety and small batch" production mode, Chinese SMEs should adopt such strategies as changing employee values, reshaping corporate culture, developing complex human resources, building internal efficiency and information management systems.

Intelligent Scheduling Method Oriented to Multi-Varieties and Small-Batch Production Mode

In terms of characteristics of scheduling problem in multi-varieties and small-batch production mode, three additional constraint conditions including combination constraint, priority scheduling and machine load balance are supplemented based on classical Job-shop scheduling model in order to reflect the actual scheduling activities better. An improved genetic annealing algorithm is proposed aiming at the above model, and the requirement of supplemented constraint conditions is satisfied by means of improving the decoding process of chromosomes. The key technologies of algorithm are discussed in detail. The feasibility and validity of the proposed model and algorithm is demonstrated through simulated computation.

Implementing CIMS in Shenyang Blower Works

Computer‐integrated manufacturing (CIM) has been paid serious attention in most of the industrialized countries. Shenyang Blower Works (SBW), the largest enterprise developing, designing and fabricating turbine compressors, blowers and large fans in China, made its production plan according to orders due to its single or small batch production mode. For this SBW decided to adopt the CIM philosophy and build CIM systems (CIMS) to increase the enterprise’s market competitiveness. SBW has begun to develop CIMS going through the steps of feasibility study, primary design, detailed design and a partial implementation‐breakthrough project since 1990. A relatively perfect computer support environment for the computer‐integrated manufacturing system in Shenyang Blower Works (SB‐CIMS), especially for the primary integrated application software systems, has been completed. This accounted for various economic benefits.