International Journal of Industrial Engineering & Production Research

---

The main advantage of heuristic or metaheuristic algorithms compared to exact optimization methods is their ability in handling large-scale instances within a reasonable time, albeit at the expense of losing a guarantee for achieving the optimal solution. Therefore, metaheuristic techniques are appropriate choices for solving NP-hard problems to near optimality. Since the parameters of heuristic and metaheuristic algorithms have a great influence on their effectiveness and efficiency, parameter tuning and calibration has gained importance. In this paper a new approach for robust parameter tuning of heuristics and metaheuristics is proposed, which is based on a combination of Design of Experiments (DOE), Signal to Noise (S/N) ratio, Shannon entropy, and VIKOR methods, which not only considers the solution quality or the number of fitness function evaluations, but also aims to minimize the running time. In order to evaluate the performance of the suggested approach, a computational analysis has been performed on the Simulated Annealing (SA) and Genetic Algorithms (GA) methods, which have been successfully applied in solving respectively the n-queens and the Uncapacitated Single Allocation Hub Location combinatorial problems. Extensive experimental results showed that by using the presented approach the average number of iterations and the average running time of the SA were respectively improved 12 and 10.2 times compared to the un-tuned SA. Also, the quality of certain solutions was improved in the tuned GA, while the average running time was 2.5 times faster compared to the un-tuned GA.

---

Location-allocation of facilities in service systems is an essential factor of their performance. One of the considerable situations which less addressed in the relevant literature is to balance service among customers in addition to minimize location-allocation costs. This is an important issue, especially in the public sector. Reviewing the recent researches in this field shows that most of them allocated demand customer to the closest facility. While, using probability rules to predict customer behavior when they select the desired facility is more appropriate. In this research, equitable facility location problem based on the gravity rule was investigated. The objective function has been defined as a combination of balancing and cost minimization, keeping in mind some system constraints. To estimate demand volume among facilities, utility function(attraction function) added to model as one constraint. The research problem is modeled as one mixed integer linear programming. Due to the model complexity, two heuristic and genetic algorithms have been developed and compared by exact solutions of small dimension problems. The results of numerical examples show the heuristic approach effectiveness with good-quality solutions in reasonable run time.

---

The quality of manufactured products is characterized by many controllable quality factors. These factors should be optimized to reach high quality products. In this paper we try to find the controllable factors levels with minimum deviation from the target and with a least variation. To solve the problem a simple aggregation function is used to aggregate the multiple responses functions then an imperialist competitive algorithm is used to find the best level of each controllable variable. Moreover the problem has been better analyzed by Pareto optimal solution to release the aggregation function. Then the proposed multiple response imperialist competitive algorithm (MRICA) has been compared with Multiple objective Genetic Algorithm. The experimental results show efficiency of the proposed approach in both aggregation and non aggregation methods in optimization of the nonlinear multi-response programming.

---

In this paper the problem of serial batch scheduling in a two-stage hybrid flow shop environment with minimizing Makesapn is studied. In serial batching it is assumed that jobs in a batch are processed serially, and their completion time is defined to be equal to the finishing time of the last job in the batch. The analysis and implementation of the prohibited transference of jobs among the machines of stage one in serial batch is the main contribution of this study. Machine set-up and ready time for all jobs are assumed to be zero and no Preemption is allowed. Machines may not breakdown but at times they may be idle. As the problem is NP-hard, a genetic algorithm is developed to give near optimal solutions. Since this problem has not been studied previously, therefore, a lower bound is developed for evaluating the performance of the proposed GA. Many test problems have been solved using GA and results compared with lower bound. Results showed GA can obtain a near optimal solution for small, median and large size problems in reasonable time.

---

A three stage production system is considered in this paper. There are two stages to fabricate and ready the parts and an assembly stage to assembly the parts and complete the products in this system. Suppose that a number of products of different kinds are ordered. Each product is assembled with a set of several parts. At first the parts are produced in the first stage with parallel machines and then they are controlled and ready in the second stage and finally the parts are assembled in an assembly stage to produce the products. Two objective functions are considered that are: (1) to minimizing the completion time of all products (makespan), and (2) minimizing the sum of earliness and tardiness of all products (∑_i▒(E_i∕T_i ) . Since this type of problem is NP-hard, a new multi-objective algorithm is designed for searching locally Pareto-optimal frontier for the problem. To validate the performance of the proposed algorithm, in terms of solution quality and diversity level, various test problems are made and the reliability of the proposed algorithm, based on some comparison metrics, is compared with two prominent multi-objective genetic algorithms, i.e. NSGA-II and SPEA-II. The computational results show that performance of the proposed algorithms is good in both efficiency and effectiveness criterions.

---

The paper is devoted to solution of some problems in nuclear power station generating unit intellectual control systems using genetic algorithms on the basis of control system model development, optimizations methods of their direct quality indices and improved integral quadratic estimates. Some mathematical vector models were obtained for control system multicriterion quality indices with due consideration of stability and quality indices criteria, this increasing the reliability of optimal control system synthesis. Optimal control systems with fuzzy controllers were synthesized for nuclear reactor, steam generator and steam turbine, thus allowing comparison between fuzzy controllers and traditional PID controllers. Mathematical models built for nuclear power station generating unit control systems, including nuclear reactor, steam generator, steam turbine and their control systems interacting under normal operational modes, which permitted to perform parametrical synthesis of system and to study various power unit control laws. On the basis of power unit control system models controllers were synthesized for normal operational modes.

---

There are many real problems in which multiple responses should be optimized simultaneously by setting of process variables. One of the common approaches for optimization of multi-response problems is desirability function. In most real cases, there is a correlation structure between responses so ignoring the correlation may lead to mistake results. Hence, in this paper a robust approach based on desirability function is extended to optimize multiple correlated responses. Main contribution of the current study is the synthesis of ideas considering correlation structure in robust optimization through defining joint confidence interval and desirability function method. A genetic algorithm was employed to solve the introduced problem. Effectiveness of the proposed method is illustrated through some computational examples and some comparisons with previous methods were performed to show applicability of the proposed approach. Also, a sensitivity analysis was provided to show relationship of correlation and robustness in these approaches.

---

Abstract One of the basic assumptions in hub covering problems is considering the covering radius as an exogenous parameter which cannot be controlled by the decision maker. Practically and in many real world cases with a negligible increase in costs, to increase the covering radii, it is possible to save the costs of establishing additional hub nodes. Change in problem parameters during the planning horizon is one of the key factors causing the results of theoretical models to be impractical in real world situations. To dissolve this problem in this paper a mathematical model for dynamic single allocation hub covering problem is proposed in which the covering radius of hub nodes is one of the decision variables. Also Due to NP-Hardness of the problem and huge computational time required to solve the problem optimally an effective genetic algorithm with dynamic operators is proposed afterwards. Computational results show the satisfying performance of the proposed genetic algorithm in achieving satisfactory results in a reasonable time. Keywords: hub location problem, dynamic hub covering problem, flexible covering radius, dynamic genetic algorithm.

---

In this paper, we propose a vibration damping optimization algorithm to solve a fuzzy mathematical model for the single-item capacitated lot-sizing problem. At first, a fuzzy mathematical model for the single-item capacitated lot-sizing problem is presented. The possibility approach is chosen to convert the fuzzy mathematical model to crisp mathematical model. The obtained crisp model is in the form of mixed integer linear programming (MILP) which can be solved by existing solver in crisp environment to find optimal solution. Due to the complexity and NP-hardness of the problem, a vibration damping optimization (VDO) is used to solve the model for large-scale problems.  To verify the performance of the proposed algorithm, we computationally compared the results obtained by the VDO algorithm with the results of the branch-and-bound method and two other well-known meta-heuristic algorithms namely simulated annealing (SA) and genetic algorithm (GA). Additionally, Taguchi method is used to calibrate the parameters of the meta-heuristic algorithms. Computational results on a set of randomly generated instances show that the VDO algorithm compared with the other algorithms can obtain appropriate solutions.

---

A teaching-learning-based optimization (TLBO) algorithm is a new population-based algorithm applied in some applications in the literature successfully. Moreover, a genetic algorithm (GA) is a popular tool employed widely in many disciplines of engineering. In this paper, a hybrid GA-TLBO algorithm is proposed for the capacitated three-stage supply chain network design (SCND) problem. The SCND problem as a strategic level decision-making problem in supply chain management is an NP-hard class of computational complexity. To escape infeasible solutions emerged in the problem of interest due to realistic constraints, combination of a random key and priority-base encoding scheme is also used. To assess the quality of the proposed hybrid GA-TLBO algorithm, some numerical examples are conducted. Then, the results are compared with the GA, TLBO, differential evolution (DE) and branch-and -bound algorithms. Finally, the conclusion is provided.

---

In this article, a multi objective model is presented to select and allocate the order to suppliers in uncertainty condition and in a multi source, multi customer and multiproduct case in a multi period state at two levels of supply chain. Objective functions considered in this study as the measures to evaluate suppliers are cost including purchase, transportation and ordering costs, timely delivering, shipment quality or wastages which are amongst major quality aspects, partial and general coverage of suppliers in respect of distance and finally suppliers weights making the products orders amount more realistic. The major limitations are price discount for products by suppliers which are calculated using signal function. In addition, suppliers weights in the fifth objective function is calculated using fuzzy Topsis technique. Lateness and wastes parameters in this model are considered as uncertain and random triangular fuzzy number. Finally the multi objective model is solved using two multi objective algorithms of Non-dominated Sorting Genetic Algorithm (NSGA-II) and Particle Swarm Optimization (PSO) and the results are analyzed using quantitative criteria Taguchi technique was used to regulate the parameters of two algorithms. 

---

In this paper a novel modelling and solving method has been developed to address the so-called resource constrained project scheduling problem (RCPSP) where project tasks have multiple modes and also the preemption of activities are allowed. To solve this NP-hard problem, a new general optimization via simulation (OvS) approach has been developed which is the main contribution of the current research. In this approach, the mathematical model of the main problem is relaxed and solved then the optimum solutions were used in the corresponding simulation model to produce several random feasible solutions for the main problem. Finally, the most promising solutions were selected as the initial population of a genetic Algorithm (GA). To test the efficiency of the problem, several test problems were solved by the proposed approach and according to the results, the proposed concept has a very good performance to solve such a complex combinatoral problem. Also, the concept could be easily applied for other similar combinatorics. 

---

In this research a jobshop scheduling problem with an assembly stage is studied. The objective function is to find a schedule which minimizes completion time for all products. At first, a linear model is introduced to express the problem. Then, in order to confirm the accuracy of the model and to explore the efficiency of the algorithms, the model is solved by GAMS. Since the job shop scheduling problem with an assembly stage is considered as a NP-hard problem, a hybrid algorithm is used to solve the problem in medium to large sizes in reasonable amount of time. This algorithm is based on genetic algorithm and parallel variable neighborhood search. The results of the proposed algorithm are compared with the result of genetic algorithm. Computational results showed that for small problems, both HGAPVNS and GA have approximately the same performance. And in medium to large problems HGAPVNS outperforms GA.

---

In the new strategies of purchasing and production, suppliers play a key role in achieving of competitive capability for big companies. The selection of suitable suppliers is a critical component of this strategy. The problem of allocation order to supplier is a multi-objective problem that includes fuzzy parameters and also suppliers usually consider discount due to different levels of purchase amount. Since there is not a multi-objective fuzzy model for allocating orders in the literature that consider discount and shortfall simultaneously in the planning horizon of multi-products, this research proposes a new model includes minimization of costs, delays and the percentage of defective parts as objective functions. Jimenez method is used to defuzzy price, demand, delay and percentage of defective parts. Since the model is NP-hard, the two metaheuristic algorithms, NSGAII and MOPSO that their parameters were set by the Taguchi method has been developed. According to the results of numerical problems, the proposed algorithms can provide a good approximation of the optimal solutions for intended objectives.
 

---

In this paper, an inventory-routing problem for a network of appliance repair service is discussed including several repair depots and customers. The customer in this network makes a demand to have his/her faulty appliance repaired. Then, the repairman is assigned to the demand based on the skill needed for repairing of appliance differing for each one. The assigned repairman picks up the faulty appliance from the customer place using the vehicle for transferring faulty appliances to repair depot. The vehicle for picking up and delivering the appliances has a maximum capacity. Additionally, the repair depot needs spare parts to repair the faulty appliances that is supplied either by the supplier or lateral transshipment from the other depots. The capacitated vehicle inventory-routing problem with simultaneous pickup and delivery is NP-hard which needs special optimization procedure. Regarding the skill of repairman, it becomes more complex. Many solution approaches have been provided so far which have their pros and cons to deal with. In this study, an augmented angle-based sweep method is developed to cluster nodes for solving the problem. Finally, the heuristic is used in the main body of genetic algorithm with special representation.

---

It is indispensable that any manufacturing system is consistent with potential changes such as fluctuations in demand. The uncertainty also makes it more essential. Routing Flexibility (RF) is one of the necessities to any modern manufacturing system such as Flexible Manufacturing System (FMS). This paper suggests three mixed integer nonlinear programming models for the Unequal–Area Stochastic Dynamic Facility Layout Problems (UA–SDFLPs) by considering the Routing Flexibility. The models are proposed when the independent demands follow the random variable with the Poisson, Exponential, and Normal distributions. To validation of the proposed models, many small-sized test problems has solved that derived from a real case in literature. The large-sized test problems are solved by the Genetic Algorithm (GA) at a reasonable computational time. The obtained results indicate that the discussed models for the UA–SDFLPs are valid and the managers can take these models to the manufacturing floor to adapt to the potential changes in today's competitive market.
 

---

Project scheduling is one of the most important and applicable concepts of project management. Many project-oriented companies and organizations apply variable cost reduction strategies in project implementation. Considering the current business environments, in addition to lowering their costs, many companies seek to prevent project delays. This paper presents a multi-objective fuzzy mathematical model for the problem of project scheduling with the limitation of multi-skilled resources able to change skills levels, optimizing project scheduling policy and skills recruitment. Given the multi objectivity of the model, the goal programming approach was used, and an equivalent single-objective model was obtained. Since the multi-skilled project scheduling is among the NP-Hard problems and the proposed problem is its extended state, so it is also an NP-Hard problem. Therefore, NSGA II and MOCS meta-heuristic algorithms were used to solve the large-sized model proposed using MATLAB software. The results show that the multi-objective genetic algorithm performs better than the multi-objective Cuckoo Search in the criteria of goal solution distance, spacing, and maximum performance enhancement.

---

Nowadays, supply chain management (SCM) is an interesting problem that has attracted the attention of many researchers. Transportation network design is one of the most important fields of SCM. In this paper, a logistics network design is considered to optimize the total cost and increase the network stability and resiliency. First, a mixed integer nonlinear programming model (MINLP) is formulated to minimize the transportation time and transportation cost of products. The proposed model consists of two main stages.
One is a normal stage that minimizes the transportation and holding costs, all manufacturers are also assumed to be healthy and in service. In this stage, the quantity of customer demand met by each manufacturer is eventually determined.
The second is the resilience stage. A method is presented by creating an information network in this supply chain for achieving the resilient and sustainable production and distribution chain that, if some manufacturers break down or stop production, Using the Restarting and load sharing scenarios in the reactive approach to increase resilience with accepting the costs associated with it in the supply network and return to the original state in the shortest possible time, the consequences of accidental failure and shutdown of production units are managed.
Two capacities are also provided for each manufacturer

• Normal capacity to meet the producer's own demand
• Load sharing capacity, Determine the empty capacity and increase the capacity of alternative units to meet the out-of-service units demand

In order to solve the model, we used GAMS & Matlab software to find the optimal solutions. A hybrid priority-based Non-dominated Sorting Genetic Algorithms (NSGA-II) and Sub-population Genetic Algorithm (SPGA- II) is provided in two phases to find the optimal solutions. The solutions are represented with a priority matrix and an Allocated vector. To compare the efficiency of two algorithms several criteria are used such as NPS, CS and HV. Several Sample problems are generated and solved that show the Sub-population Genetic Algorithm (SPGA- II) can find good solutions in a reasonable time limit.

---

---

This paper presents a model for minimizing the makespan in the flow shop scheduling problem. Due to the impact of increased workloads, flow shops are becoming more popular and widely used in industries. To solve the challenge of minimizing makespan, a Hybrid-Heuristic-Metaheuristic-Genetic-Algorithm (HHMGA) is proposed. The proposed HHMGA algorithm is tested using the simulation software and demonstrated with steel industry data. The results are compared with those of the best available flow shop problem algorithms such as Palmer’s slope index, Campbell-Dudek-Smith (CDS), Nawaz-Enscore-Ham (NEH), genetic algorithm (GA) and particle swarm optimization (PSO). According to empirical results and relative differences from the lower bound, the proposed technique outperforms the three heuristics and two metaheuristics algorithms in three of six cases, while the remaining three produce the same results as the NEH heuristic. In comparison to the steel industry's regular job scheduling technique, the simulation model based on HHMGA can save 4642 hours. It was discovered that the suggested model enhanced the job sequence based on the makespan requirements.