A robust optimal scheduling system based on multi-performance driving for complex manufacturing systems

Question

1. What factors affect production scheduling?

2. How is preventive maintenance integrated in production scheduling?

3. What is robust scheduling?

4. What is closed-loop optimal scheduling?

Accepted Answer

Production scheduling of complex manufacturing systems involves allocating limited resources according to time series to achieve target values. Factors such as tasks, resources, and task completion time play a crucial role. However, traditional scheduling theory assumes fixed parameters, while actual production processes face random disturbances like equipment failures, order changes, and worker differences. These uncertainties necessitate adaptive scheduling schemes with robustness to ensure successful completion and desired outcomes. The research focuses on multi-objective driven robust optimal scheduling, considering dynamic dispatching, equipment maintenance, and robustness measures like cycle time, on-time delivery rate, equipment unavailability, and robustness measure. The proposed scheduling system integrates intelligent manufacturing dynamic operation optimization theory to solve the adaptive closed-loop optimal scheduling problem in manufacturing systems.

Accepted Answer

Preventive maintenance is integrated into production scheduling by considering it as a decision variable or constraint. It is studied in workshop management and production scheduling. Integrated optimization of production scheduling and equipment maintenance is divided into deterministic and uncertain tasks. In deterministic problems, sufficient preventive maintenance is arranged to ensure high equipment reliability. In uncertain problems, preventive maintenance is necessary along with considering possible machine failures. Various strategies have been adopted, such as scheduling preventive maintenance when reliability is reduced to a threshold value, using a fixed period strategy, and considering multi-objective evolutionary algorithms. The impact of equipment failures on workpiece processing is also considered, and the reliability of equipment is improved through preventive maintenance. Integrated optimization models have been established for equipment preventive maintenance and production scheduling, particularly in single-machine systems and flow shops. However, the existing literature primarily focuses on single-machine systems and flow shops, and there is a need for integrated scheduling optimization of complex manufacturing systems under uncertain random factors.

Accepted Answer

Robust scheduling consists of defining the robustness of scheduling, specifying the robustness index, and the method for measuring it. It involves selecting the appropriate robust scheduling method based on problem characteristics and available information. Scheduling robustness is measured by comparing actual scheduling goals with initial goals, using maximum regret value, or expected average weighted tardiness. Dual objective Pareto optimization methods optimize performance and robustness under random machine failures. Robust scheduling aims to maintain good performance in the presence of uncertain factors and dynamic processing environments.

Accepted Answer

Closed-loop optimal scheduling is a data-driven approach that addresses the challenges of universality, adaptability, and robustness in manufacturing scheduling. It involves a simulation system to assist in decision-making optimization of production scheduling. Feng et al. proposed a production scheduling optimization method based on an intelligent factory multi-level simulation system. They established an MILP optimization model and integrated simulation system for scheduling optimization and simulation modules. Closed-loop optimization mechanism was introduced to improve the robustness of the scheduling optimization model. Yu et al. proposed a self-organized scheduling method that adapts to the characteristics of the semiconductor production line. Qiao et al. designed a combined scheduling rule using the response surface method to optimize weight parameters. The closed-loop dynamic scheduling method effectively guides production operations of complex manufacturing systems, considering scheduling optimization, control integration, model consistency, and manufacturing system performance.

Accepted Answer

MLP, a typical artificial neural network, plays a significant role in deep learning. It is widely used for remote sensing, prediction of meteorological and natural data, and various applications such as disease prediction, environment detection, chemical material analysis, and intelligent system control. Theoretically, MLP can approach arbitrary nonlinear functions even with complex data distributions, given deep enough layers and sufficient single perceptions in the hidden layer. However, a network too deep may lead to overfitting, necessitating optimization methods for consistent problem-solving. The whale optimization algorithm, a population-based meta-heuristic algorithm, has been successfully applied to optimize MLP parameters, enhancing its performance in model prediction and parameter optimization.

Accepted Answer

The proposed joint decision-making method integrates production scheduling and equipment maintenance. It utilizes the multi-objective optimization method based on IWOA-MLP to build an integrated optimization system. This system aims to optimize the CT ODR, EA, and RM of scheduling schemes. The scheduling parameters are adaptively adjusted through closed-loop optimization, allowing the scheduling rules to adapt to current working conditions. This approach effectively improves production efficiency, maintains equipment availability, avoids deterioration of scheduling performance, and ensures the robustness of the scheduling scheme.

Accepted Answer

The study assumes that setting time, processing preparation time, and transportation time are not considered. Equipment maintenance and processing cannot occur simultaneously. Workpieces follow sequential processes with multiple machines. Equipment faults are maintainable, allowing continuous processing post-repair. Interrupted tasks resume after maintenance. No inherent machine priority exists. Specific process routes exist between operations of the same job. These assumptions help streamline the scheduling process and optimize efficiency in the study's context.

Accepted Answer

The scheduling problem solved in this research has the following constraints: EQUATION Formula (1) indicates that the job cannot be disassembled during processing; Formula (2) indicates that each job can only be processed by one machine at the same time; Formula (3) represents process constraints; Formula (4) represents machine constraints; Formula (5) indicates that the process can only be started after the equipment completes the process; Formula (6) indicates that the completion time of the equipment is constrained by the processing time of the process on the corresponding machine; Constraint (7) ensures that the waiting time of the workpiece between two adjacent processes cannot exceed the upper limit. These constraints are essential in determining the feasibility and efficiency of the scheduling solution.

Accepted Answer

The multi-performance driven closed-loop adaptive optimal scheduling framework aims to address the challenges in complex manufacturing systems by generating a new scheduling scheme to improve performance indicators. It consists of four modules: semiconductor wafer fabrication line simulation system, sample generation module, off-line training module, and online scheduling module. The framework utilizes data related to equipment maintenance and production status to output scheduling parameter combinations, driven by CT, ODR, EA, and RM. This approach ensures robustness and the ability to absorb and cope with uncertain disturbances, ultimately optimizing the performance index of the manufacturing system.

Accepted Answer

The closed-loop optimal scheduling system consists of four main parts: 1) combined scheduling rules integrating equipment maintenance and job dispatching, 2) feature selection using an improved immune algorithm, 3) analysis of current production status and expected performance indicators through IWOA-MLP algorithm, and 4) optimization of scheduling parameters using IWOA-MLP algorithm to achieve multi-objective driven adaptive optimization process. These parts work together to create a dynamic and efficient scheduling system that adapts to changing production conditions and optimizes performance indicators.

Accepted Answer

The proposed semiconductor production line scheduling rule comprehensively considers process constraints and equipment maintenance by considering multiple factors related to process constraints and flexible maintenance of equipment. It evaluates downstream equipment load, equipment maintenance urgency, job delivery urgency, and equipment availability. The rule follows a step-by-step process, starting with checking if the workpiece meets the process constraint. If it does, the rule calculates the process constraint urgency and delivery urgency. It then determines the dispatching priority based on customer priority, equipment maintenance period, emergency degree of equipment maintenance, load degree of downstream equipment, and equipment unavailability. The rule normalizes the dispatching priority and equipment maintenance priority to make informed decisions on workpiece dispatching and equipment maintenance scheduling.

Accepted Answer

The multi-layer perceptron model based on improved whale optimization algorithm training is a research-based approach that utilizes the whale optimization algorithm to optimize the training process of a multi-layer perceptron model. The algorithm is inspired by the bubble net foraging behavior of humpback whales and consists of two main stages: shrink and surround, and position update. The shrink and surround stage involves whales sensing and surrounding their prey, with the best search agent being defined as the target prey. The position update stage includes spiral position update and random search methods to explore and update whale positions. The improved whale optimization algorithm (IWOA) addresses the limitations of the traditional whale optimization algorithm by introducing chaotic Tent mapping for initial population position, a new nonlinear parameter, fitness control mechanism, Harris Hawk siege mechanism, and Gaussian detection position control. This enhanced algorithm aims to accelerate convergence speed, adapt to complex nonlinear problems, prevent stagnation, and improve optimization accuracy.

Accepted Answer

Multi-layer perceptron (MLP) is a neural network structure composed of input, hidden, and output layers. Each layer consists of multiple neurons, and the bottom layer is the input layer, the middle layer is the hidden layer, and the last layer is the output layer. The input layer neurons receive information, while the hidden layer neurons process the input information through fully connected connections. The output layer neurons are responsible for the computer's cognition of the input information. The output of the output layer is a multi-category logical regression. The parameters of the MLP model, including connection weight matrix and offset vector, are optimized using algorithms like the Improved Whale Optimization Algorithm (IWOA).

Accepted Answer

Simulation experiments can enhance the effectiveness of multiobjective flexible job shop scheduling models by providing a platform to test and verify the constructed scheduling model and solution strategy. In the case of a semiconductor manufacturing enterprise in Shanghai, a simulation platform can be used to replicate the complex manufacturing system with large scale, complex process flow, and high reentry. By simulating the production lines, processing methods, equipment, processing zones, products, and processes flow, researchers can evaluate the performance of the scheduling model under realistic conditions. This allows for the identification of potential improvements and optimization strategies, ultimately leading to more efficient and robust scheduling solutions. Additionally, simulation experiments can help in understanding the impact of various factors on the scheduling model's effectiveness, such as changes in production volume, equipment availability, and process variations. Overall, simulation experiments provide a valuable tool for researchers to validate and refine multiobjective flexible job shop scheduling models, ensuring their practical applicability in real-world manufacturing environments.

Accepted Answer

The IWOA-MLP algorithm's closed-loop optimization is compared through simulation experiments and analysis of multiple dimensions. Driving sources such as CT, ODR, EA, and RM are selected to evaluate the effectiveness of the algorithm. The comparison includes the average CT of workpieces, ODR, EA, and RM of scheduling schemes. The algorithm's performance is assessed against self-organizing and predictive scheduling systems proposed in previous works. This comprehensive analysis helps determine the algorithm's efficiency and potential improvements in the scheduling system.

Accepted Answer

The three aspects analyzed in the experimental results section are 'combined scheduling rules', 'closed-loop optimization', and 'scheduling system'. These aspects are crucial in understanding the efficiency and effectiveness of the manufacturing system. 'Combined scheduling rules' refer to the integration of various scheduling techniques to optimize production processes. 'Closed-loop optimization' involves the continuous improvement of the system by using feedback from the production process. 'Scheduling system' focuses on the implementation and performance of the scheduling algorithms. By analyzing these aspects, researchers can identify areas for improvement and enhance the overall productivity of the manufacturing system.

Accepted Answer

The ratio of the WIP quantity in oxidation area to that in the whole manufacturing system is calculated by dividing the WIP quantity in the oxidation area by the total WIP quantity in the manufacturing system. This ratio helps in understanding the distribution of work-in-progress (WIP) across different areas of the manufacturing system. A higher ratio indicates a higher concentration of WIP in the oxidation area, which may suggest inefficiencies or bottlenecks in that specific area. By analyzing this ratio, researchers can identify areas that require optimization or improvement to enhance overall manufacturing efficiency.

Accepted Answer

The closed-loop optimal scheduling system based on IWOA-MLP algorithm outperforms other systems in terms of CT, ODR, EA, and RM performance. Compared to the combined scheduling rules, it improves CT by 5.357%, ODR by 2.319%, EA by 8.491%, and RM by 9.537%. When compared to the combined scheduling rules, the system based on machining cycle drive improves CT by 7.872%, ODR by 1.912%, EA by 3.066%, and RM by 5.450%. The system based on ODR drive significantly improves ODR performance, while the EA driven system reduces EA performance. The RM driven system enhances RM performance. Overall, the closed-loop optimal scheduling system based on IWOA-MLP algorithm demonstrates superior performance in various aspects compared to other scheduling systems.

Accepted Answer

The proposed IWOA-MLP-based scheduling method enhances semiconductor manufacturing system performance by integrating equipment maintenance, process constraints, and real-time status correlation. It utilizes a dynamic dispatching rule and constructs a performance prediction model based on 11 processing statuses. The model outputs optimized dispatching parameters for each status, allowing the system to adaptively acclimatize to new production environments. Simulation results demonstrate that this method outperforms conventional scheduling policies in average cycle time, equipment availability, on-time delivery rate, and robustness measure. Additionally, it self-adaptively satisfies the dynamic environment of the manufacturing system, leading to overall performance improvement.

A robust optimal scheduling system based on multi-performance driving for complex manufacturing systems

Chat with Paper

AI Agents for this Paper

Most frequently asked questions

1. What factors affect production scheduling?

2. How is preventive maintenance integrated in production scheduling?

3. What is robust scheduling?

4. What is closed-loop optimal scheduling?

5. How does MLP contribute to deep learning?

6. What is the proposed joint decision-making method for production scheduling?

7. What assumptions are made in the scheduling problems of this study?

8. What are the constraints in the scheduling problem solved in this research?

9. What is the purpose of a multi-performance driven closed-loop adaptive optimal scheduling framework?

10. What are the four main parts of the closed-loop optimal scheduling system?

11. How does the proposed semiconductor production line scheduling rule consider process constraints and equipment maintenance?

12. What is the multi-layer perceptron model based on improved whale optimization algorithm training?

13. How does multi-layer perceptron work?

14. How can simulation experiments enhance the effectiveness of multiobjective flexible job shop scheduling models?

15. How does the IWOA-MLP algorithm's closed-loop optimization compare?

16. What are the three aspects analyzed in the experimental results section?

17. What is the ratio of WIP in oxidation area?

18. How does the closed-loop optimal scheduling system based on IWOA-MLP algorithm compare to other systems?

19. How does the proposed IWOA-MLP-based scheduling method improve semiconductor manufacturing system performance?

References

Robustness measures and robust scheduling for job shops

A WOA-Based Optimization Approach for Task Scheduling in Cloud Computing Systems

A Knowledge-Based Cooperative Algorithm for Energy-Efficient Scheduling of Distributed Flow-Shop

Predictable scheduling of a single machine with breakdowns and sensitive jobs

Multi-objective genetic algorithm for energy-efficient hybrid flow shop scheduling with lot streaming

Related Papers (5)

Optimal preventive maintenance scheduling in semiconductor manufacturing

The case for near real-time production scheduling in a highly automated semiconductor manufacturing environment

Modelling and scheduling preventative maintenance in semiconductor manufacturing industry with MAs

Daily planning and scheduling system for the EDS process in a semiconductor manufacturing facility.

Dynamic scheduling strategy with efficient node availability prediction for handling divisible loads in multi-cloud systems