Zhonghua Han

Multiple rules decision-based DE solution for the earliness-tardiness case of hybrid flow-shop scheduling problem

The earliness/tardiness (E/T) case of hybrid flow-shop scheduling problem (HFSP) is an NP hard problem, which is difficult to deal with; however, the existence of the multi-rules relating to the practical production increases the complexity of this problem. How to solve the combinatorial optimisation problem effectively and optimally is still an open issue today. In this paper, the joint scheduling strategy of differential evolution (DE) algorithm and factor space-based multiple rules decision method is used to solve this E/T scheduling problem. Firstly, DE algorithm is used to make global assignment and obtain each job’s process route. Secondly, factor space method is used to describe the scheduling rules in production process; then a scheduling decision method based on variable weight comprehensive function is considered to figure out the jobs’ operating priority in buffer area during the local production assignment between stages; subsequently the starting time of each job can be determined. Finally, u...

A study and analysis of digital image processing and recognition algorithms

In this paper, we study and analyse the recognition methods for digital image. In the digital image preprocessing, the conservative smoothing, mean filtering, Gaussian sharpening and binarisation are used together so as to guarantee the effectiveness of the digital feature extraction. We propose an edge-tracking method, a distance feature information method and a feature information extraction method for recognising digital images. And, those three methods are evaluated and compared by experimental tests.

Cost Optimization Problem of Hybrid Flow-shop Based on PSO Algorithm

A PSO-algorithm-based job scheduling method that takes production cost as optimization object is presented in this paper. The cost optimization model of HFSP, in which production cost is considered as an optimal factor, is constructed. PSO is used to take global optimization, make the production task assignment and find which machine the jobs should be assigned at each stage, which is also called the process route of the job. After that the local assignment rules are used to determine the job’s starting time and processing sequence at each stage. The total production cost converted by time-based scheduling results is comprehensively considering the processing cost, waiting costs, and the products storage costs. The numerical results show the effectiveness of the algorithm after comparing between multi-group programs.

Solving the Two-Objective Shop Scheduling Problem in MTO Manufacturing Systems by a Novel Genetic Algorithm

With the characteristics of un-touching, high-automatic and high-speed, we measured 3D shape by digital moiré patterns. And it can measure 3D shape without compensator, auxiliary mirror, hologram and other assistant modules. So it is widely used in measurement of 3D shape. But with the negative effect of grating area and CCD resolution of the camera, it has some problems in large aperture, steep aspheric surface, even off-axis aspheric testing. Focus on these problems; this paper established a sub-region splicing measurement of aspheric surface by using the combination of digital moire patterns and digital phase shifting technology. This paper also illustrated the basic theory, and the mathematical implementation procedure. In all, sub-region splicing digital moiré patterns can be considered as another method, which beyond compensation tests to measurement of aspheric surface.

Time-window-based combined objective DE algorithm for hybrid flow-shop

In this paper, a time-window-based hybrid flow-shop scheduling model is presented for manufacturing enterprise for solving production scheduling optimisation problem in the rolling production process. Time window deviation sum and the earliness/tardiness (E/T) penalty sum’s optimisation goal are modelled, and differential evolution (DE) global optimisation algorithm is used to conduct the global optimisation. The assigned tasks to determine a job’s machining path in the time window frame can be finished in each stage. The hybrid flow-shop scheduling optimisation method based on time window can improve the equipment utilisation and reduce scheduling conflict situations. Our time-window-based hybrid flow-shop scheduling optimisation method is evaluated by multiple sets of programs’ numerical analysis, and the results show that the proposed approach can meet the requirement of business rolling production scheduling very well.

Research on Equipment Fault Diagnosis Method Based on Multi-Sensor Data Fusion

In this paper, we presented an equipment fault diagnosis method based on multi-sensor data fusion, in order to solve the problems such as uncertainty, imprecision and low reliability caused by using a single sensor to diagnose the equipment faults. We used a variety of sensors to collect the data for diagnosed objects and fused the data by using D-S evidence theory, according to the change of confidence and uncertainty, diagnosed whether the faults happened. Experimental results show that, the D-S evidence theory algorithm can reduce the uncertainty of the results of fault diagnosis, improved diagnostic accuracy and reliability, and compared with the fault diagnosis using a single sensor, this method has a better effect.

The Controller Design for DC Motor Speed System Based on Improved Fuzzy PID

In this paper, a controller is designed based on improved fuzzy PID to solve the problem that the dc motor performance of speed and dynamic is poor when using the conventional PID controller for the lack of adaptive capacity of the controller parameters. The improved fuzzy control algorithm is used for the tuning of PID controller to get good speed performances, which automatically adjust the parameter of PID controller according to the motor speed. The simulation results show that the improved fuzzy PID control with the advantages of fast response, small overshoot and strong anti-interference capability can effectively improve the dynamic characteristics and steady state accuracy.

DE solution for the earliness/tardiness case of Hybrid Flow-shop Scheduling problem with priority strategy

The earliness/tardiness (E/T) case of Hybrid Flow-shop Scheduling problem (HFSP) is an NP hard problem, which is difficult to deal with, however, the local assignment existing in the practical production increases the complexity of this problem. How to solve the combinatorial optimization problem effectively and optimally is still an open issue today. In this paper, differential evolution algorithm (DE) combined with priority strategy is used to solve this E/T scheduling problem. Firstly, DE algorithm is used to make global assignment and obtain each jobs process route. Secondly, the operating priority of the jobs in buffer area deduced from the expectation completion sequence is used to direct the local production assignment between stages, then the starting time of each job can be determined. Finally, under the constraints of the due-date, the global optimization with the minimal penalty sum of E/T is obtained. Several scheme comparisons with simulation results show the effectiveness of the proposed method.

Improved NSGA-II algorithm for multi-objective scheduling problem in hybrid flow shop

In this paper, multi-objective optimization for hybrid flow shop scheduling problem has been studied. The delivery time penalty and the load imbalance penalty are taken as the evaluation metrics. We describe the optimization framework for this hybrid flow shop problem, and design an improved NSGA-II algorithm for solution searching. Specifically, a multi-objective dynamic adaptive differential evolution algorithm (MODADE) is proposed to enhance the searching efficiency of the general differential evolution operations. MODADE calculates the similarity between different individuals based on their Hamming distance, and dynamically generates the high-similarity individuals for the population. We compare MODADE compared with the state-of-the-art algorithms, and the numerical result shows that the proposed MODADE algorithm outperforms others in terms of the algorithm convergence, the number and distribution of Pareto solutions.

Development for granular computing-based multi-agent system for data fusion process

In data fusion systems, the characteristics of the information from the sensors include diversity, complexity and uncertainty. In this paper, the data fusion of the granular computing-based multi-layer structure is studied. Neural network and fuzzy system are adopted for the inference mechanism to construct an equivalent fuzzy logic system. Neural network clustering is used to cluster the concept lattices with different formal contexts. And in each concept lattice, fuzzy clustering is used to cluster the formal contexts. The design of the data fusion middleware in the multi-agent system MAS enables the two-step data fusion. This design is used to solve the issues caused by the imprecise, incomplete, fuzzy or contradictory inference. Simulation results on the fire detection in the intelligent building environment show the effectiveness and the feasibility of this design.