Improving Renewable Energy Recovery Efficiency in Variable Pressure Source Systems Through BP Neural Network Optimization

The variable pressure source system is widely used in renewable energy recovery scenarios. However, the instability of the new energy input and the nonlinearity of the power generation system can lead to problems such as more difficult tracking and control of the maximum power point and poor power generation quality. To address the problem of lower efficiency under varying inputs, which is prevalent in renewable energy generation, this paper establishes an energy recovery system model based on power conversion rectifier topology, designs a speed current double closed-loop control strategy, proposes an online variable-step maximum efficiency point tracking method based on experience curves, and tests the overall energy recovery effect through system simulation and comparative experiments. The simulation results show that the maximum efficiency point tracking method proposed in this paper reduces the number of optimization searches by 50% and improves optimization speed. The experiment results show that, under the drastic changes of the input, the method proposed in this paper could reduce the total harmonic distortion to 5.65%, improve the energy recovery efficiency by 10.976%, and reduce the fluctuation ratio of the voltage to 2.43%. This study can provide an important reference for the collection and utilization of new energy sources.