The Efficiency Of The Rectifier Unit To Optimize The Research And Application Of Genetic Algorithms

At present technology level, the smelting of the nonferrous metals is a great power-consumpting process. So the technology of energy-saving in smelting industry will directly bring the great economic benefits.The smeltery of Zhuzhou is a large-scale smelting enterprise of nonferrous melals of our country. The technicians in this factory pointed out the defaults of the existing mode of power supply:1.In the policy-making level,the managers make decisions on load distribution just by experiences.It's difficult for them to find the optimal distribution even they know the optimal distribution exists.2. In the communication layer , the RS232 serial port is widely used ,butthere are some problems : low rate of communication speed,no intelligence and it's difficult to extend (he communication service.3. In the control layer , the speed and the accuracy ol control can be improved .so , the managers of this factory have put forward the request for technological transformation, which is described in the following:1. In the policy-making level they adopted the advanced technology to determine the configuration rectifiers and the distribution of load, on the premise of a satisfying power factor and aimed at the lowest power wasting.2. In the communication layer ,the technology of fieldbus should be used to construct a industry network ,so the information can be tranmissioned between the policy-making level and the control layer.3. In the control layer advanced equipments and algorithm should be used to improve the speed and accuracy of control.In this paper ,the float-encoded genetic algorithm is improved to optimize the efficiency of the coordinate high-power rectification unit. To deal with the restrictions we give different penalty cocfs to different populations. In selection, we preserved the elitist. In cross and mutaion ,we adopt the adaptive rate to overcome the problem of premature. The result shows that this algorithm is convergent to the global optimum . The speed of convergence can meet the requirement of the engineering. This method is easier than the engineering method ,and the result is more precision.