Heuristics with novel approaches for cyclical multiple parallel machine scheduling in sugarcane unloading systems

Variable Neighborhood Strategy Adaptive Search to Solve Parallel-Machine Scheduling to Minimize Energy Consumption While Considering Job Priority and Control Makespan

“…The proposed HDE algorithm, a special encoding and decoding scheme, was constructed to deal with the machine-eligibility constraint, and chaos theory was applied to determine the parameter settings of the DE algorithm. Kusoncum et al [17] presented the traditional DE with embedded heuristics to obtain nearoptimal solutions in realistically sized machine scheduling, including capacitated machine restrictions and sequencing-independent setup-time considerations. Results revealed that the proposed method's performance tended to find a new optimal solution during the simulation, while the local-search-based heuristics were trapped at some local optima, and the DE was insufficient for search intensification.…”

“…Kusoncum et al [17] introduced another application of VaNSAS in a sugar mill as a computational tool for scheduling sugarcane-vehicle-unloading systems. The objective was to minimize the makespan of parallel-machine-capacity scheduling with a cyclic sequence, and machine restriction included sequencing-independent setup time.…”

“…In the production process, the electricity consumption of each job is sometimes different, indicating that an operation with high electricity consumption is arranged for low processing time jobs, and an operation with low electricity consumption is arranged for high processing time jobs, so that the electricity cost can be decreased [5,6,13]. Paying too much attention to energy cost as an outstanding optimization objective may cause a long makespan and imbalanced workload on machines, as well as lead to bottlenecks, late deliveries, and even untimely machine failure [14][15][16][17].…”

“…Parallel machine scheduling is defined as sequencing and assigning jobs into machines when similar types of machines are available and jobs can be scheduled in these machines. A variety of sequencing and/or processing restrictions often exist when decision makers try to minimize some related objective functions [14][15][16][17][18][19][20][21][22][23][24]; however, most enterprises still use advanced machines running alongside outdated ones. In contrast to the old ones, modern machines are usually adjusted to work at a high speed and save energy.…”

“…Therefore, this addressed scheduling is both an objective optimization problem and an NP-hard problem. The differential evolutionary (DE) algorithm is suitable to solve this kind of problem because it can obtain non-dominated solutions in a single run and has been successfully applied to optimization problems [17,20,24,[27][28][29]. In addition, metaheuristics based on local search methods, such as the variable neighborhood strategies adaptive search (VaNSAS), have been successfully applied to solve many combinatorial optimizations problems [30][31][32][33][34][35][36][37][38][39][40], which inspired us to develop a parallel-machine-scheduling model, and to propose VaNSAS and new neighborhood strategies: (1) solution destroy and repair (SDR); (2) track-transition method (TTM); and (3) multiplier factor (MF).…”