The operation and maintenance costs of offshore wind farms are more than twice as high as those onshore, and the uneven fatigue load distribution in wind turbines leads to a high maintenance frequency for wind farms, which makes the problem of high operation and maintenance costs of wind farms more prominent. Under this background, an optimal dispatching strategy for the active power of an offshore wind farm is proposed, which takes into account the fatigue load distribution in wind turbines. First, a fatigue load assessment model based on a fatigue load coefficient for offshore wind turbines is put forward, in which the relationship of fatigue load coefficient with wind turbine active power output and ambient wind turbulence fluctuation is constructed, the comprehensive fatigue damage of wind turbines is assessed through the fatigue load coefficient, and the problem of wind turbine fatigue load optimization is transformed into an active power control problem. Second, the wind farm wake loss is minimized under the minimum thrust coefficient control strategy, and the fatigue load distribution in wind turbines is balanced on the premise that the total fatigue load of the wind farm is not increased and the total power output from the wind farm is not lower than the value under the maximum power point tracking control strategy for wind turbines, so that the maintenance frequency of each unit is synchronized. Therefore, the overall maintenance frequency of the wind farm is reduced. Finally, simulation results show that the proposed method can effectively reduce the maintenance cost of one wind farm, providing technical support for the demand of cost reduction and efficiency improvement in the era of offshore wind power parity.

The Three-North region in China is rich in wind and PV resources. However, the system flexibility is insufficient due to the operation constraint of “fixing power based on heat” on the combined heat and power unit, which leads to the phenomenon of wind power abandonment especially in the winter heating period. To fully exploit the system flexibility and improve the consumption level of wind power, an optimal scheduling model for an integrated energy system with operation flexibility is proposed. First, according to the physical characteristics of different types of power supply, the indices that quantify the flexibility are introduced by combing factors such as load, volatility and randomness, and a supply and demand balance model for flexibility is also established. Then, the flexibility coupling mechanism on both sides of source and load is analyzed, and the space of wind power consumption after the system flexibility is improved is quantified. On this basis, the influence of different new energy penetration rates on the system flexibility is considered. Finally, the result of a simulation example proves that the proposed scheduling model can coordinate with the flexibility resources on both sides of source and load to achieve mutual gain, which verifies its feasibility in improving the flexibility.

The goal of avirtual power plant (VPP) is to aggregate and use a large number of distributed generations (DGs) with tiny capacities. A load frequency control (LFC) model for the secondary frequency regulation of thermal power units is constructed with the time-segmented participation of a VPP that aggregates energy storage, electric vehicles and wind turbines, and a kind of centralized automatic generation control strategy with VPP is proposed. The frequency modulation model for aggregated resources in the VPP is analyzed, the constraints on each frequency modulation link are taken into account, and the minimization of frequency modulation costs and area control erroris taken as an objective. Meanwhile, model predictive control is used to optimize the allocation of power output from DGs within the VPP. The proposed LFC model for a two-region power system is simulated using MATLAB/Simulink, and the frequency error and power deviation of the tie line are compared before and after the VPP takes part in the system frequency modulation. According to the simulation results, it is found that the multi-region LFC with participation of VPP can reduce frequency disturbances and power deviation of the tie line brought by power fluctuations, and the proposed control strategy can achieve the optimal allocation of active output. Comparison with the conventional PID regulation, this strategy also has better control outcomes.

Since the electrolytic hydrogen production (EHP) load has a large capacity, a fast response speed and flexible and adjustable power, it is a natural physical resource that can provide energy for frequency regulation. However, due to limitations on the existing two kinds of rectifiers used in EHP (i.e., thyristor and PWM rectifiers), it is difficult to realize the participation of EHP load in the system’s multi-scale frequency regulation. In this paper, a novel multi-scale frequency regulation control strategy for a hybrid rectifier used in EHP is proposed. A droop relationship between the firing angle of a thyristor rectifier in the hybrid rectifier and the system frequency is built, so as to realize the primary frequency regulation by regulating the hydrogen production power flowing through the thyristor rectifier according to changes in the system frequency. The auxiliary converter adopts virtual synchronous machine control to realize the inertia compensation, thereby smoothing the rate of change of frequency and gaining time for the primary frequency regulation performed by thethyristor rectifier. Through the coordination between the thyristor rectifier and the auxiliary converter in the hybrid rectifier, the multi-scale frequency regulation of EHP load is achieved. Finally, a low power experimental platform was established to verify the effectiveness of the proposed control method.

Hybrid electric vehicles (HEVs) have been widely developed in recent years owing to their advantages such as environmental protection, low noise and energy saving. The signal intersections are an important part of urban roads, where the traffic congestion is easy to occur when the road vehicle density is high, thus reducing the efficiency of the entire road network and affecting the fuel economy of vehicles. With the development of new technologies such as intelligent transport systems, the problem of HEV energy management considering the road traffic information is becoming a research hotspot. The research results of speed planning and energy management strategy (EMS) for HEVs considering road slopes and combination of single and mixed vehicles at signal intersections are introduced, the existing methods of speed trajectory optimization and EMS based on fusing the road traffic information which are proposed by scholars both at home and abroad are summarized, and the development of EMS considering the traffic information is also discussed.

To improve the economy and stability of microgrid operation, it is necessary to optimize the capacity of microgrid reasonably according to the meteorological information and load information. Under this background, a mathematical model of distributed power sources is established. According to the constraint conditions and operation strategy for the system, the number of distributed power sources is taken as an optimization variable, and the minimum total cost is taken as an objective function. The improved immune particle swarm optimization algorithm is used to optimize the capacity of microgrid. First, normal distribution is used for initialization, so as to increase the population diversity. Then, the nonlinear inertia factor, adaptive inertia weight and chaotic disturbance operator are used to improve the convergence speed and accuracy of the algorithm. Experimental results show that the proposed method is reasonable and can effectively reduce the investment cost, providing a reference value for the capacity optimization of microgrid.

Under partial shading conditions, a multi-peak state will appear in the P-U curve of photovoltaic array. In this case, the conventional maximum power point tracking (MPPT) algorithm is prone to fall into local extremum, which cannot accurately track the maximum power point of a photovoltaic power generation system in time. Aimed at this problem, a MPPT control strategy based on an improved dung beetle optimizer (IDBO) algorithm is proposed. First, the targeted initialization of the dung beetle population is optimized, and the Levy flight strategy is introduced in the position update process. Through simulation verification and physical experiment verification on a MATLAB/Simulink platform, the IDBO algorithm can find the global maximum power point faster and more stably than the traditional algorithm under both the static and dynamic conditions.

The state-of-health (SOH) of lithium-ion batteries (LIBs) is a key parameter that characterizes their real lifespan. To improve the accuracy of SOH estimation for LIBs, a novel SOH estimation method based on health factors and improved long short-term memory (LSTM) neural network is proposed. First, ten health factors that are highly correlated with the changes in SOH are selected as the characteristics of SOH. Second, the particle swarm optimization (PSO) algorithm is introduced to optimize the hyper-parameter of the LSTM neural network, and an estimation model based on PSO-LSTM is constructed. Finally, the extracted health factors are used as model input to achieve an accurate estimation of SOH. Results show that when 70%, 60% and 50% of data from the whole sample are taken as the training set respectively and the remaining data is taken as the test set, the SOH estimation errors are mainly distributed in a range of less than 2%. Meanwhile, the PSO-LSTM model demonstrates higher estimation accuracy and better robustness compared with the existing LSTM and feed-forward neural networks (FNN).

The state-of-health (SOH) of a battery is difficult to be directly measured, so an accurate estimation of SOH is critical for ensuring its safe operation. Aimed at the problem of SOH estimation of lithium-ion batteries, an SOH estimation method based on improved sparrow search algorithm-Gaussian process regression (ISSA-GPR) is proposed. First, ISSA is used to optimize the parameters of a GPR model, and the GPR model based on ISSA is constructed. Then, the capacity increment curve is analyzed to extract the health factors that characterize the battery capacity degradation as the input of the model, the optimal voltage interval length with the peak as the center area is determined by ISSA, and the SOH of batteries is obtained to perform estimation. Finally, the public experimental data sets are applied to verify the proposed method. Experimental results show that the proposed ISSA-GPR method can accurately estimate the SOH of batteries.

The wireless power transfer (WPT) technology, which can provide flexible and convenient power supply for unmanned aerial vehicles (UAVs) without human assistance, is an important research direction for the intelligent development of UAVs in the future. To meet the practical requirements of lightweight, constant current (CC) and constant voltage (CV) charging, and a high-power transmission efficiency for a UAV-WPT system, a high-efficiency CC/CV output regulating method for the UAV-WPT system based on voltage-controlled capacitor (VCC) is proposed. On the transmitter side, a VCC is utilized as the compensated capacitor. By dynamically adjusting the equivalent impedance of the VCC, the output current and voltage from the system can be effectively controlled. Meanwhile, zero-voltage switching can also be achieved in the inverter in a wide load range without any extra auxiliary power sources or inductors. The operating principle for the proposed UAV-WPT system and its power loss model are analyzed in detail. In addition, a DC bias voltage regulation module, a closed-loop control strategy and a parameter design method for the VCC are designed. Experimental results show that under the rated output conditions of 25.2 V and 6 A, the system’s overall efficiency can reach 88.8%.