Research On Photovoltaic Array Monitoring And Mismatch Detection Based On Wireless Sensor Network

Solar photovoltaic(PV)power generation has a huge installed capacity in the world because of its advantages of environmental protection,noiselessness,and a wide range of sources.However,due to the long-term operation in complex and harsh outdoor environments,PV arrays are prone to various failures,which will reduce the power generation efficiency and safe operation.Therefore,it is particularly important to monitor the PV arrays.There are some defects in current monitoring systems.For example,wired systems suffers complicated wiring and high costs;most systems only monitor the PV array or PV string parameters without PV module level information and the capability to automatically detect,locate,and alert the faults.To address these shortcomings,based on two-level wireless sensor network(WSN),a low-cost on-line monitoring system is proposed in this paper to achieve module-level monitoring and automatic detection and location of mismatch faults.The monitoring system consists of four parts:sensing nodes installed on PV modules(PVM),sensing and routing nodes of PV sub-arrays(PVA)installed at combiner boxes,mismatch detection and location algorithm and a data management software(DMS)running on a host computer.The PVM node consists of a microcontroller MSP430G2553,a power management module,a nRF24L01 wireless transceiver,PVM voltage acquisition circuit and a temperature sensor.PV module and lithium polymer battery are applied to ensure the power supply of PVM node in different weathers.A precise resistor voltage divider scheme is used to collect the module voltage,and a temperature sensor is utilized to collect the backplane temperature of PVM.Then the acquired parameters are transmitted to the corresponding PVA node via the first-level nRF24L01 star network.The PVA node consists of a microcontroller MSP430F149,power management chips,a nRF24L01 and ZigBee wireless module,a temperature sensor,a Hall voltage sensor and some Hall current sensors.The PVA node receives corresponding module parameters from PVM nodes through the first-level nRF24L01 network,and collects PV array voltage,PV string current,ambient irradiance and ambient temperature.Then,it transmits all the data to DMS through the second-level ZigBee mesh network.Based on the PV string current and module voltage,an improved Hampel identifier algorithm is proposed to automatically detect and locate the mismatch faults of PV module and PV strings.Firstly,a PV array simulation model is built using the MATLAB/Simulink,and several typical mismatch faults simulation are carried out to obtain faulty data to verify the feasibility of the proposed algorithm.Secondly,the algorithm is further verified experimentally based on the measured voltage and current data.Both the simulation and the measured results indicate that the designed algorithm can accurately locate the mismatched PV strings and modules,additionally,it overcome the false alarm of the basic Hampel identifier.Finally,a DMS is designed using MATLAB/GUIDE and MySQL database to store,display and query the monitored data,and realize automatic mismatch fault detection,positioning and alarm.Experimental results show that the proposed system can monitor the electrical and environmental parameters of PV modules and arrays under different operating condition in real time,and the monitored data can be used to locate the mismatched PV strings and modules according to the expected design goals.The work of this paper can provide some ideas for the online monitoring and mismatch detection of PV arrays.