配电系统中多目标条件下可控负荷的最优控制

发布时间：2019-04-27 08:35

【摘要】：电力需求侧管理（demand side management, DSM）是指通过采取有效的措施来引导、激励或辅助电力用户改变用电习惯，提高供用电效率，以降低负荷费用，平滑负荷曲线，减少网络损耗、提高供电可靠性等为目的而采取的一项对环境、电力公司、电力用户及社会都有巨大意义的工程。作为需求侧管理最重要的方面，负荷管理从本质上改变了以往纯粹依靠增加发电侧发电机组容量来应对负荷快速增长的局面，充分调动了电力用户参与电网安全稳定合理运行与发展的积极性，电力公司与用户协作来提高电力系统运行稳定性并减少双方的供电、用电成本，最大化双方利益。本文研究负荷管理的重要组成部分--可控负荷在配网中的多目标优化控制策略，并将改进的多目标粒子群优化算法(improved multi-objectiveparticle swarm optimization, IMOPSO)应用到了可控负荷模型的优化中，相关仿真波形验证了该可控负荷多目标控制策略在降低配网网损与用户电费方面的有效性和可靠性。论文针对现阶段配网中存在的各种问题，如负载率过低，高峰时段峰值较大但持续时间一般较短，系统发电容量不足以满足快速增长的负荷，可再生能源大规模接入带来的波动性功率输出及负荷随机性等造成的频率、电压波动甚至崩溃等，提出了利用对配网中不同种类可控负荷进行控制的解决方案，并阐述了可控负荷在解决这些问题时的优越性。论文重点对空调、热水器、冰箱、电动汽车等典型可控负荷的工作特性进行了研究，根据“黑盒子”理论只考虑其外部工作特性，建立了简化的可控负荷数学模型。过程的简化与参数的概率选择提高了模型在实际求解中的简易性及准确性，在此基础上论文提出了基于低压配网的可控负荷多目标优化控制策略，该策略利用启发式算法对不同节点、不同种类、不同时段的可控负荷工作状态进行优化，以期减小配网网损、降低峰谷差、减少用户电费等，控制策略灵活且不需对现有配网进行大幅改造，便于实际应用推广。针对可控负荷多目标优化问题的多维度、多约束性，论文采用多目标粒子群算法作为可控负荷控制策略的底层算法，并针对现有多目标粒子群算法存在的，如优秀个体选取方式不明确、对约束处理不够灵活等问题，提出了基于归一化函数值位与约束惩罚位的粒子比较策略，该比较策略综合考虑了粒子对应各目标函数的函数值与对各约束的违反程度，可以更好地反映粒子的适应度并引导粒子向最优解前沿加速搜索，基于相关改进设计了新的多目标粒子群算法IMOPSO，该算法可以更好地求解可控负荷多目标优化问题，给出更合理的可控负荷控制序列。论文的最后利用Matlab软件编制了相应的IMOPSO源程序，并分别以网损与电费最小为目标函数，仿真了论文提出的可控负荷多目标优化控制策略在IEEE14节点中，不同天气情况下的实际优化效果，，仿真结果证明该控制策略在不同天气情况下,均能在满足用户用电舒适度的前提下，大大降低网损与电费，具有很好的应用前景。根据蒙特卡洛法，得出了控制策略优化结果对各电动汽车渗透率及可控负荷控制比的灵敏度，为进一步研究可控负荷应用提供了方向。
[Abstract]:The power demand side management (DSM) refers to the use of effective measures to guide, motivate or assist the power user to change the power utilization habit, improve the power supply efficiency, to reduce the load cost, to smooth the load curve, to reduce network loss, A project for the purpose of improving the reliability of power supply, etc., is of great significance to the environment, the power company, the power user and the society. As the most important aspect of the demand side management, the load management changes the situation that the load is rapidly growing by increasing the capacity of the power generation side generating set in essence, and fully realizes the initiative of the power user to participate in the safe and stable operation and development of the power grid, The utility and the user cooperate to improve the operation stability of the power system and reduce the power supply, the power utilization cost and the benefit of the two parties. In this paper, the important component of load management--the multi-objective optimization control strategy in the distribution network is studied, and the improved multi-objective particle swarm optimization (IMPSO) is applied to the optimization of the controllable load model. The related simulation waveforms verify the effectiveness and reliability of the controllable load multi-target control strategy in reducing the network loss and the user's electricity fee. In view of the problems existing in the distribution network at present, such as the low load rate, the high peak period, the duration is generally shorter, and the power generation capacity of the system is not enough to meet the negative of the rapid growth The frequency, voltage fluctuation and even breakdown caused by the large-scale access of the renewable energy source and the large-scale access of the renewable energy source and the like, and the solution of controlling the different kinds of controllable loads in the distribution network are put forward. The case, and expounds the superiority of the controllable load in the solution of these problems The paper focuses on the working characteristics of typical controllable loads such as air-conditioning, water heater, refrigerator and electric vehicle. According to the "black box" theory, only the external working characteristics are considered, and the simplified controllable load number is established. The simplified and parameter selection of the process improves the simplicity and accuracy of the model in the practical solution. On the basis of this, the paper presents a multi-objective optimization control strategy based on the control load of the low-voltage distribution network. The controllable load working state of the same kind and different time period is optimized, with the aim of reducing the network loss, reducing the peak-to-valley difference, reducing the user's electric charge and the like, In this paper, the multi-objective particle swarm optimization algorithm is used as the bottom-layer algorithm of the controllable load control strategy, and the multi-objective particle swarm algorithm is used as the bottom-layer algorithm of the controllable load control strategy, and the multi-objective particle swarm optimization algorithm is applied to the existing multi-objective particle swarm optimization algorithm, such as the excellent individual selection. In this paper, we put forward a particle comparison strategy based on the normalized function value bit and the constraint penalty bit, which takes into account the function value of the particle's corresponding target function and the constraint on each constraint. The invention can better reflect the fitness of the particles and guide the particles to accelerate the search to the optimal solution front, and design a new multi-target particle swarm optimization algorithm (IMEPSO) based on the related improvement, which can better solve the problem of multi-objective optimization of the controllable load and give a more reasonable and controllable negative effect. In the end of this paper, the corresponding IMOPSO source program is developed by using the Matlab software, and the net loss and the electric charge are the least objective function, and the control strategy of the multi-objective optimization of the controlled load is simulated in the IEEE14 node under different weather conditions. The simulation results show that the control strategy can greatly reduce the net loss and the electric charge under different weather conditions, and has the advantages of greatly reducing net loss and electric charge, According to the Monte-Carlo method, the sensitivity of the control strategy optimization result to the control ratio of the permeability and the controllable load of each electric vehicle is obtained, and the controllable load can be further studied.
【学位授予单位】：湖南大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TM921.5

【参考文献】