光伏电池板所处环境的非线性变化使得光伏电池的功率保持在最大功率点(maximum power point，MPP)非常困难。传统的最大功率点跟踪(maximum power point tracking，MPPT)方法普遍存在技术缺陷，无法满足当前需求。针对光伏发电MPPT问题，该文提出了一种基于麻雀搜索算法优化的极限学习机(sparrow search algorithm-extreme learning machine，SSA-ELM)神经网络控制器的MPPT方法。与传统技术相比，该MPPT方法在稳定性、速度、超调和MPP的振荡等方面的效果均较好。使用MATLAB/Simulink平台进行仿真实验，验证了所提控制策略及理论分析的正确性。
Photovoltaic MPPT method based on SSA-ELM neural network controller
[Objective]Working at the maximum power point of photovoltaic cells can improve energy utilization efficiency and ensure sufficient power supply.However,most distributed photovoltaic power generation systems are located in natural environments,and the nonlinear changes in the environment in which photovoltaic panels operate make it very difficult to maintain their maximum power point.Traditional maximum power point tracking methods,such as perturbation observation,conductance increment,and particle swarm optimization,yield unsatisfactory results because of their different shortcomings.[Methods]To address the problem of maximum power point tracking in photovoltaics,this paper proposes a maximum power point tracking(MPPT)method for photovoltaics through research on neural network technology and is based on an extreme learning machine(ELM)neural network controller optimized using the sparrow search algorithm(SSA).Herein,a training database is established by analyzing and collecting data on the maximum power curve of photovoltaic cells at different temperatures and light intensities.The proposed neural network controller is trained to form a two-input and one-output controller.The voltage value corresponding to the maximum power point is calculated in real time based on the collected temperature and light intensity data and transmitted to the subsequent boost converter circuit controller.Further,a modulated wave is generated through PWM modulation,which controls the output voltage of the boost converter circuit to track the maximum power point voltage and builds a simulation experimental platform using MATLAB/Simulink.[Results]The simulation results show the following:① the proposed method has a high calculation speed and can rapidly track the reference voltage value,making the voltage value accurate.② The proposed method directly controls the voltage changes without oscillation or overshoot,resulting in a stable system.③ The proposed method shows good search accuracy,convergence speed,and stability.The performance of the other three aforementioned mainstream tracking methods and the output voltage of the proposed method under sudden increase or decrease in light intensity are analyzed.[Conclusions]The results indicate that the proposed method can rapidly achieve maximum power point tracking without oscillation and overshoot issues.Compared with the traditional methods,the proposed method performs better in terms of search accuracy,convergence speed,and stability.
photovoltaic cellsmaximum power point trackingsparrow search algorithmextreme learning machine
万方数据
维普
