Comprehensively considering the operation cost and safety constraints of nuclear power, an optimal operation scheme of large‐scale nuclear power plant participating in peak load regulation of power system is proposed. After quantitatively analysing the peak load regulation cost of nuclear power, the optimal objective is set to minimise the total operation cost including the fuel cost, the start–stop cost, and the peak load regulation cost. On the basis of the normal constraints of power scheduling, safety constraints of the nuclear power unit's daily output performance and power characteristic constraints of the pumped storage unit are considered. The mixed integer programming method is used to achieve the solution of the optimal operation scheme. Simulation results from a regional power system show that the proposed optimal operation scheme can effectively reflect the real characteristic of nuclear power and apply to power dispatching analysis under the background of the large‐scale access of nuclear power. In comparison to other two typical schemes of peak load regulation, the proposed optimal operation scheme of peak load regulation is more economic and efficient.

The effective prediction of bus load can provide an important basis for power system dispatching and planning and energy consumption to promote environmental sustainable development. A bus load forecasting method based on variational modal decomposition (VMD) and bidirectional long short-term memory (Bi-LSTM) network was proposed in this article. Firstly, the bus load series was decomposed into a group of relatively stable subsequence components by VMD to reduce the interaction between different trend information. Then, a time series prediction model based on Bi-LSTM was constructed for each sub sequence, and Bayesian theory was used to optimize the sub sequence-related hyperparameters and judge whether the sequence uses Bi-LSTM to improve the prediction accuracy of a single model. Finally, the bus load prediction value was obtained by superimposing the prediction results of each subsequence. The example results show that compared with the traditional prediction algorithm, the proposed method can better track the change trend of bus load, and has higher prediction accuracy and stability.

In this study, a Pelton turbine and governor system dynamic model with deflector control was developed. The Pelton turbine established reasonable parameters for the deflector control model to effectively restrain unit frequency rise, which played the role of a similar thermal power overspeed protection. Since a Pelton turbine typically contains multiple needles, the concept of average needle opening was employed in the model. An improved orthogonal learning biogeography‐based optimisation (IOLBBO) algorithm, which adopted good point set to initialise habitat, elites to maintain the tactics, and an orthogonal learning strategy to improve global search capabilities, was developed. The parameters were identified with the IOLBBO algorithm based on measured data. To develop the model in this study, various types of Pelton turbine and governor system models with different refinement degrees were analysed. Large fluctuations, like a load rejection process, verified the validity of the established deflector control model. Additionally, digital deflector control characteristics were achieved by computer programming. The non‐linear Pelton turbine and water diversion system model with an elastic water hammer model that considered a non‐linear relationship between the average needle opening and mechanical power developed in this study demonstrated superior simulation results in power system stability analysis.