Design Optimization of a New Hot Heat Sink with a Rectangular Fin Array for Thermoelectric Dehumidifiers

The objective of this research was aimed at conducting an experimental investigation to study the heat sink performance of a new rectangular fins array. Various operating conditions were considered, namely the distance between the fan and the fins, varied from 0 mm to 40 mm, heat supplied to the sink and the fan voltage. It was concluded that a fan installed at 30 mm height from the fins is recommended as the hot side temperature was the lowest. Next a pre-experimentation of small scale prototype of thermoelectric Dehumidifier (TED) was designed and constructed. It was composed of two thermoelectric (TE) cooling modules, MT2-1, 6-127, (two in serial) mounted between the rectangular fin heat exchangers with dimension 140 × 240 × 35 mm. The hot side was cooled by a ventilation fan whereas the air flow on the cold side was free convection. The effect of position of fan was investigation experimentally. Preliminary tests confirmed the good performance of the hot heat sink design for the intended thermoelectric application.     

Multiple colony ant algorithm for job-shop scheduling problem

Ant colony optimization (ACO) is a metaheuristic that takes inspiration from the foraging behaviour of a real ant colony to solve the optimization problem. This paper presents a multiple colony ant algorithm to solve the Job-shop Scheduling Problem with the objective that minimizes the makespan. In a multiple colony ant algorithm, ants cooperate to find good solutions by exchanging information among colonies which are stored in a master pheromone matrix that serves the role of global memory. The exploration of the search space in each colony is guided by different heuristic information. Several specific features are introduced in the algorithm in order to improve the efficiency of the search. Among others is the local search method by which the ant can fine-tune their neighbourhood solutions. The proposed algorithm is tested over set of benchmark problems and the computational results demonstrate that the multiple colony ant algorithm performs well on the benchmark problems.