基于人工智能算法的电力系统可靠性评估模型及应用研究

Research on the reliability evaluation model and application of power systems based on artificial intelligence algorithms

针对电力系统可靠性评估精确度提升问题,提出一种基于PSO算法优化的深度信念网络可靠性评估模型。通过深度信念网络,提高了模型的学习深度,在引入PSO算法对参数优化后,进一步提升了模型对电力系统的可靠性评估质量。实验表明,所提PSO-DBN算法,相较于传统DBN算法,训练收敛速度更快,且样本重构效果更好,在对电力系统的评估测试中,RMSE值最终稳定在0.1以下,更接近真实值,评估精度提高明显;相较于Monte Carlo法和PSO-BP模型的评估结果,所提PSO-DBN模型虽然用时较PSO-BP模型稍长,但评估精度大幅增长,RMSE值分别降低了94.95%和81.98%,模型质量更好,值得在电力系统可靠性评估中进一步研究和推广。

A reliability evaluation model of deep belief network based on PSO algorithm is proposed to improve the accuracy of automatic evaluation of reliable power system.Through deep belief network,the learning depth of the model is improved.After introducing PSO algorithm to optimize the parameters,the reliability evaluation quality of the model on the power system is further improved.Experiments show that compared with the traditional DBN algorithm,the proposed PSO-DBN algorithm has faster training convergence speed and better sample reconstruction effect.In the evaluation test of power system,the RMSE value finally stabilizes below 0.1,which is closer to the real value,and the evaluation accuracy is improved significantly.Compared with the evaluation results of Monte Carlo method and PSO-BP model,although the proposed PSO-DBN model takes a little longer time than PSO-BP model,the evaluation accuracy is greatly increased,and the RMSE values are reduced by 94.95% and 81.98%,respectively.The model quality is better,and it is worthy of further study and promotion in the reliability evaluation of power system.