【作者】 秦明明；

【导师】 王坚；

【作者基本信息】 东华大学 ， 检测技术与自动化装置， 2010， 硕士

【摘要】 电力工业的根本任务是以安全为中心,在充分合理地利用能源和运行设备能力的条件下,保证安全经济发、供电,以满足国民经济各部门的电能需求。电力系统供应着现代化社会生产和生活的大部分能量,相应地,也消耗着大量的一次能源——煤、石油等。对于电力这样重要的能源转换系统,提高其运行效率、实现其运行优化的必要性是显而易见的。电力市场的运行目标是:在满足系统安全稳定运行的条件下,促进发电厂的竞争,以发电成本,网损,辅助服务等方面成本之和最低为优化目标,根据机组报价,确定发电计划,实时调度各个发电公司的机组发电,以满足用电负荷要求。粒子群优化(Partical Swarm Optimization-PSO)算法是美国心理学家Kennedy和电气工程师Eberhart受鸟类觅食行为的启发而提出的一种基于群体智能理论的新兴演化计算技术。这种算法以其实现容易、精度高、收敛快等优点引起了学术界的重视,并且在解决实际问题中展示了其优越性.目前已广泛应用于函数优化,神经网络训练,模糊系统控制以及其他遗传算法的应用领域。由于标准PSO算法在优化过程中表现出来一定的问题:(1)参数控制:对不同的问题,如何选择合适的参数来达到最优效果。(2)缺乏速度的动态调节:爬山能力不强,有时在达到一定的精度后,很难再找到更好的解。(3)早熟:粒子群过早收敛,使寻优停滞。(4)后期收敛慢。因此我们需要对标准粒子群算法进行改进,以使其达到最优的结果。本文提出了一种改进的的粒子群算法,采用对算法参数的改进的方法,对粒子群算法的惯性权重和加速常数都进行了改进,为了使目标函数尽快达到最优,本文的粒子群算法对粒子的飞行速度也进行了限制。并将其应用于优化函数,和标准粒子群算法、遗传算法进行了比较,结果证明此方法很有效。又把改进的粒子群算法用于电力系统的有功优化问题中,针对电力系统有功优化调度,提出了一种改进的粒子群算法,该算法考虑了火电厂的煤耗量,污染物排放量,以及线路损耗等,通过分别求解各个单目标优化问题和定义各单项目标的隶属度函数,把多目标优化问题转化为单目标优化问题,从整体上降低电力系统的发电成本。该算法以标准粒子群算法为基础,对其参数进行了改进,并对其搜索速度加以限制。将其应用于电力系统的3机组模型,算例仿真结果表明该算法节省了收敛时间,具有收敛速度快,计算精度高的优点。最后将改进的粒子群算法用于焦作电网的模拟优化调度,节省了发电成本,对焦作电网目前的状况进行了改善。更多还原

【Abstract】 The fundamental task of the electric power industry is taking the security as a center,in an adequate and rational use of energy and ability to run the equipment,under the conditions to ensure safe and economic electricity generation.Regarding such important energy transformation of the electric power system,enhances the necessity of its operating efficiency, realizes its movement optimizes is obvious.The targets of power market operations are:promote competition between power plans, optimize costs,make daily plans,save energy,protect environment etc.on the basis of safe and stable power system operation.Particle swarm optimization(Partical Swarm Optimization-PSO) algorithm is a new theory based on swarm intelligence evolution of computing technology.This algorithm is proposed by American psychologist Kennedy and Electronics Engineers,Eberhart who were inspired by the foraging behavior of birds.Its easy implementation,high accuracy and fast convergence advantages attracted academic attention,and to solve real problems demonstrated its superiority. Currently,it has been widely applied to function optimization,neural network training,fuzzy system control,and other genetic algorithm applications.Since the standard PSO algorithm in the optimization process shown by certain questions:(1) Parameter control:a range of issues,how to select the appropriate parameters to achieve optimal results.(2) the lack of dynamic adjustment of the speed:climbing capability is not strong, sometimes up to a certain accuracy,it’s hard to find a better solution.(3) Early:Particle Swarm premature convergence,so that optimization stagnation.(4) The late slow convergence.Therefore, we need to improve the standard particle swarm algorithm in order to reach the best results.This paper presents an improved particle swarm algorithm,using the method of improving algorithm parameters,the particle swarm optimization algorithm inertia weight and acceleration constants has been improved.The algorithm was applied to optimize the function,and compared to the standard particle swarm optimization algorithm,genetic algorithms,results prove this method is very effective.Then the improved particle swarm algorithm was applied to active power system optimization problems,Active for optimal scheduling of power system,this paper presents an improved particle swarm algorithm,which takes into account the volume of coal consumption of thermal power plants,pollutant emissions,as well as Lines loss.The multi-objective optimal problem can be transformed into single-objective optimal problem by means of respectively solving the various single-objective optimal problems and the definition of the objective of the individual membership function.The cost of power generation is minimized on the whole.This algorithm is based on the standard particle swarm optimization,has improved its parameter and limited the speed of its search.Apply it to 3 unit models of the power system,simulation results show that the algorithm has saved the time of searching,has fast convergence and the advantages of high accuracy.Finally,the improved particle swarm algorithm is used for simulation optimal scheduling of Jiaozuo grid,the cost of power generation on the Jiaozuo grid has been saved:the current situation has been improved.更多还原