改进蚁群算法在可靠多电源配电网规划模型中的应用

Application of Improved Ant Colony Algorithm in Reliable Multi-Source Distribution Network Planning Model

针对传统电力系统存在的故障诊断能力差、数据冗余等问题,提出一种基于自组织和深度强化学习相结合的新一代分布式电源系统可靠性优化方法。首先,利用蚁群算法对传统的单电源配电网进行建模分析;其次,采用改进蚁群算法对该模型进行训练与测试;最后,将结果作为输入,构建出一个全新的多电源配电网综合评价体系并验证了所提方法的有效性。研究表明,相比于传统的单电源配电网而言,该模型能够显著提升网络的鲁棒性、抗干扰能力以及预测准确率。同时,由于引入了自组织机制来改善网络结构,使得网络具有较高的泛化性及适应度,因此可以更好地实现对复杂电网运行状态下的动态响应,从而进一步提高预测精度和稳定性。此外,通过实验证明所提出的方法能够有效提高供电系统整体的供电效率。

In view of the problems existing in the traditional power system,such as poor fault diagnosis ability,data redundancy and so on,a new reliability optimization method for distributed power system based on self-organization and deep reinforcement learning is proposed.Firstly,the ant colony algorithm is used to model the traditional singlesource distribution network,secondly,the improved ant colony algorithm is used to train and test the model,and finally,the results are used as input,a new comprehensive evaluation system of multi-source distribution network is constructed and the effectiveness of the proposed method is verified.The results show that the model can significantly improve the robustness,anti-interference ability and prediction accuracy of the network compared with the traditional single-source distribution network.At the same time,the“self-organization”mechanism is introduced to improve the network structure,which makes the network have higher generalization and adaptability,so it can better realize the dynamic response to the operation of complex power grid,therefore,the prediction accuracy and stability can be further improved.In addition,the experimental results show that the proposed method can effectively improve the overall efficiency of power supply system.