International Journal of Industrial Engineering & Production Research

---

Redundancy technique is known as a way to enhance the reliability and availability of non-reparable systems, but for repairable systems, another factor is getting prominent called as the number of maintenance resources. In this study, availability optimization of series-parallel systems is modelled by using Markovian process by which the number of maintenance resources is located into the objective model under constraints such as cost, weight, and volume. Due to complexity of the model as nonlinear programming , solving the model by commercial softwares is not possible, and a simple heuristic method called as simulated annealing is applied. Our main contribution in this study is related to the development of a new availability model considering a new decision variable called as the number of maintenance resources. A numerical simulation is solved and the results are shown to demonstrate the effecienct of the method.

---

In single machine scheduling problems with availability constraints, machines are not available for one or more periods of time. In this paper, we consider a single machine scheduling problem with flexible and periodic availability constraints. In this problem, the maximum continuous working time for each machine increases in a stepwise manner with two different values allowed. Also, the duration of unavailability for each period depends on the maximum continuous working time of the machine in that same period, again with two different values allowed. The objective is to minimize the number of tardy jobs. In the first stage, the complexity of the problem is investigated and a binary integer programming model, a heuristic algorithm and a branch-and-bound algorithm are proposed in a second stage. Computational results of solving 1680 sample problems indicate that the branch-and-bound algorithm is capable of not only solving problems of up to 20 jobs but also of optimally solving 94.76% of the total number of problems. Based on numerical results obtained, a mean average error of 2% is obtained for the heuristic algorithm.

---

In this paper, one of the most important criterion in public services quality named availability is evaluated by using artificial neural network (ANN). In addition, the availability values are predicted for future periods by using exponential weighted moving average (EWMA) scheme and some time series models (TSM) including autoregressive (AR), moving average (MA) and autoregressive moving average (ARMA). Results based on comparative studies between four methods based on ANN and by considering the several conditions for the effective parameters in ANN show that, the generalized regression method is the best method for predicting the availability. Furthermore, results of the EWMA and three mentioned TSM are also show the better performance of MA model for predicting the availability values in future periods. 

---

 Sequence dependent set-up times scheduling problems (SDSTs), availability constraint and transportation times are interesting and important issues in production management, which are often addressed separately. In this paper, the SDSTs job shop scheduling problem with position-based learning effects, job-dependent transportation times and multiple preventive maintenance activities is studied. Due to learning effects, jobs processing times are not fixed during plan horizon and each machine has predetermined number of preventive maintenance activities. A novel mixed integer linear programming model is proposed to formulate the problem for minimizing Make Span. Owing to the high complexity of the problem; we applied Grey Wolf Optimizer (GWO) and Invasive Weed Optimizer (IWO) to find nearly optimal solutions for medium and large instances. Finally, the computational Results are provided for evaluating the performance and effectiveness of the proposed solution approaches.