Research On Islanding Detection Of Grid Connected Photovoltaic Inverter Based On DSP

With the rapid development of solar photovoltaic power generation,islanding has become an important problem that threatens the power grid,equipment safety and personnel safety.When the grid fails or for other reasons need to stop power supply,island detection has become an important means to avoid islanding.At present,the active frequency offset method and the current disturbance method in the active detection are used as the method of island detection.The system is added to the frequency offset disturbance in the active frequency offset method,additional harmonics are introduced to reduce the output power factor.Active detection method can reduce or eliminate the detection of blind spots,but in the inverter output by adding constant current disturbance will make the inverter has been underload or overload state.In this thesis,the method of island detection based on DSP is proposed,and no additional harmonics are introduced into the power grid,the effective reduction of the inverter has been in the underload or overload state,improve the life of the inverter,The thesis mainly completes the following contents.Firstly,the island detection principle and the existing detection methods in the existing problems are analyzed,the trigger minimum current disturbance method was proposed.Through the voltage transformer for low-voltage circuit breaker circuit,the arc information acquisition and processing,the signal and set the value of the comparison,the arc voltage is obtained.The detected arc voltage is sent to the DSP processor and compared with the set voltage range.If the threshold is exceeded,it is judged that the circuit breaker connected to the grid is disconnected and the minimum current disturbance amount is added.And the minimum current disturbance is obtained by collecting the voltage and current range of the common coupling point,The load current meter is calculated and the minimum current of overvoltage / undervoltage protection is determined according to the voltage range,to ensure that the load power and inverter output power balance can also be detected in the island.Secondly,the DSP TMS320F28335 as a core processor,to achieve the trigger minimum current disturbance method of the island detection.It can be real-time,quickly on the power grid signal acquisition and processing,and through the DSP to generate PWM signal control grid inverter operation,to achieve photovoltaic grid-connected inverter island detection.3kw photovoltaic grid-connected inverter system is used as an example and island detection system is designed.The voltage/current acquisition circuit in the hardware system samples the analog signal and implements the analog-to-digital conversion via the AD7606.The power supply circuit ensures that the DSP and the acquisition module are working normally.Non-isolated single-stage grid-connected structure is selected as the main circuit,the system efficiency is high,and the use of voltage source input,current source output and network control;Drive circuit to ensure reliable IGBT turn on and off,and the electrical isolation,the signal circuit and the gate circuit insulation;Using LCL filter circuit to eliminate the inverter in the process of switching the high harmonics,to avoid the impact of the grid current.ADC conversion of the digital signal is transmitted to the DSP to achieve the basic parameters of the calculation program in the system;PWM signal generated by the DSP to control the grid inverter operation,photovoltaic grid-connected inverter island detection is realized.Finally,through the Matlab environment,the correctness of the method is verified,in this foundation to build DSP as the core processor prototype,the different load conditions for island detection.The results show that the islanding is detected by this method,there is no detection blind spot,improve the inverter has been in the underload or overload state,improve the life of the inverter for the island detection method to provide the basis for research.The method proposed in this thesis provides the basis for the research of islanding detection method.