Tingxi Wang
View Full Paper
RCBA, BPNN, Power system, Economic dispatch, PSO
With the continuous expansion of the intelligence of the power system and the scale of new energy grid connection, economic dispatching, as a core technology for optimizing energy allocation, is facing three key challenges. First, the high-dimensional nonlinear constraints make the optimization solution space exhibit complex non-convex charac- teristics. The strong nonlinearity of the unit cost function and the constraint conditions leads traditional methods to easily fall into lo- cal optimum. Second, the coupling of multiple time scales increases the difficulty of cross-scale collaborative optimization. Third, the randomness of the output of new energy sources such as wind power and photovoltaic power intensifies the risk of supply and demand im- balance. These factors jointly lead to the dynamic changes of the feasible domain boundary of the optimization problem. However, the existing methods have clear quantitative limitations - the con- vergence speed drops by more than 30% in high-dimensional scenar- ios, and the prediction deviation of power generation costs exceeds 5% in high-tech energy penetration scenarios, making it difficult to meet the requirements of precise scheduling. This study proposes an economic dispatch model integrating Particle Swarm Optimization and a neural network-improved Random Chaos Bat Algorithm. Re- sults show after 98 iterations, the improved algorithm’s fitness value is stable at 103, with a median Root Mean Square Error of 0.55 and an interquartile range of 0.12. Furthermore, the fusion model evalu- ation shows after 250 iterations, the model’s power generation cost is stable at 9.48, with a 0.02 mean deviation between predicted and actual values, and the fastest single dispatch optimization time is 3.2 s. The model effectively addresses nonlinear and non-convex opti- mization challenges in power system economic dispatch, enhancing operational reliability and system economy. It also offers method- ological support for the low-carbon, intelligent dispatch of emerging power systems. College of Economics, Northwest Normal University, Lanzhou, 730070, China; e-mail: [email protected] Corresponding author: Tingxi Wang
Important Links:
Go Back
