Using NI CompactDAQ, CompactRIO, and LabVIEW to Create a Reliable, Cost Effective Energy Solution for Smart Grids

Sivaneasan Balakrishnam, Nanyang Technological University

"We successfully developed a CEMS for the NTU smart grid test bed using NI CompactDAQ, CompactRIO, and LabVIEW . The CEMS efficiently integrated the real-time solar power forecasting algorithm, improved MPPT algorithm for PV system, and robust BMS to deliver a reliable, cost-effective energy solution for smart grids."

- Sivaneasan Balakrishnam, Nanyang Technological University

The Challenge:

Creating a system for real-time forecasting of intermittent solar power and coordinated energy management.

The Solution:

Using NI CompactRIO and NI CompactDAQ hardware combined with NI LabVIEW software to develop a system that performs real-time data acquisition of solar parameters for accurate solar forecasting and coordinated energy management for efficient integration of solar photovoltaic, fuel cell, and energy stores into a smart grid.

Author(s):

Sivaneasan Balakrishnam - Nanyang Technological University
Peng Huat Cheah - Nanyang Technological University
Mong Keow Thomas Foo - Nanyang Technological University
Nandha Kumar Kandasamy - Nanyang Technological University
Kanamarlapaudi Venkata Ravi Kishore - Nanyang Technological University
Jiasheng Ren - Nanyang Technological University
Ping Lam So - Nanyang Technological University
Kuan Tak Tan - Nanyang Technological University

 

 

The Laboratory for Clean Energy Research (LaCER) at the School of Electrical & Electronic Engineering, at Nanyang Technological University (NTU) houses the award-winning microgrid energy management system prototype that incorporates software applications to manage sensing data and perform load and generation management. We extended the microgrid prototype to a smart grid prototype by incorporating home and building energy management systems, an electric vehicle charging station, and energy storage into the microgrid(Figure 1).

 

Typically, the distributed energy resources within the smart grid include solar photovoltaic (PV), wind turbine, fuel cell, and battery energy storage systems (BESS). Electronic programmable load, low voltage industrial load, and electric vehicle charging stations are the load demands in the smart grid test bed. The bidirectional power flow of the BESS in the smart grid can deliver and absorb energy based on the charging/discharging requirement. The power flow of the upstream grid is also bidirectional as the system can absorb power from the grid if the energy generated is less than the load demand, or supply it to the grid if the energy generated is more than the load demand.

 

 

We can ensure a reliable, cost-effective energy solution for the smart grid by controlling the active and reactive power injected by the DERs responsive to changes in weather conditions and dynamically regulating the power distribution based on pricing. Furthermore, we can optimize the energy used by the loads to minimize the overall costs of electricity and respond to incentives from service providers to curtail load demand at times of peak or off-peak demand. This requires a highly efficient and robust coordinated energy management system (CEMS) to schedule the power flows to the loads between grid supply and DERs using advanced control strategies with an intelligent communication interface. Figure 2 shows the block diagram of the CEMS we developed for the NTU smart grid test bed using NI LabVIEW software. The CEMS behaves as a central controller of the smart grid by sending and receiving the monitoring and control information to and from the solar forecasting module, DER inverters, and smart meters.

 

 

 

Solar Power Forecasting Using LabVIEW and NI CompactDAQ

We can improve the reliability issues caused by the intermittent nature of solar PV with an accurate solar power forecasting algorithm using real-time solar irradiance and temperature data. The energy management system can use the forecasted solar power to control the charging and discharging of the BESS to buffer the PV output (Figure 3).

 

 

The amount of solar power generated depends on the weather conditions and solar PV panel physical specifications such as efficiency, azimuth, and tilt angle. This information serves as the input to the solar power forecasting algorithm. Figure 4 shows the basic hardware setup for solar irradiance and temperature data acquisition.

 

We use the NI cDAQ-9191 Wi-Fi chassis, configured with NI Measurement & Automation Explorer software, to collect and transmit the sensor outputs of a pyranometer and a resistance thermometer device (RTD) through an NI 9215 module. The system transmits real-time data via a Wi-Fi connection to the solar power forecasting algorithm running on a remote desktop. The cDAQ-9191 ensures a stable Wi-Fi transmission at full data rate of 100 Mbps for a transmission range of 30 m in an office-like environment or up to 100 m in a line-of-sight environment.  

 

 

Figure 5 shows the results of the measured output power from the smart meter connected to an actual PV system and simulated power obtained using the developed model. The simulated output power corresponds well to the measured output power with a mean absolute percentage error (MAPE) of 3.93 percent. This verifies the accuracy of the solar power forecasting algorithm. Both the measured and simulated output power closely follow the variations of actual solar irradiance data collected from the solar sensor system.

 

 

 

Improved Maximum Power Point Tracker for PV System

PV systems have non-linear voltage and load characteristics, so the operating point of the PV system depends on the connected load. To ensure the PV system in the smart grid test bed operates at maximum possible output power, we implement an improved maximum power point tracker (MPPT) algorithm using LabVIEW. We use an NI 9225 module and an NI 9227 module to sense the PV output parameters. We applyed the current sweep and estimation perturb and perturb (EPP) algorithm to extract the maximum available power at each time instant using the current controller in the PV inverter. Figure 6 shows the real-time PV panel characteristics and MPPT output.

 

 

 

A Battery Management System for BESS Using an NI FPGA-based cRIO-9082

We prefer lithium-ion batteries compared to other rechargeable batteries because of their high power and energy densities. We must be careful not to over charge or over discharge the batteries as the overall available capacity of the series-connected battery pack is reduced due the imbalance in capacities of individual battery modules. We can bypass this by employing a battery management system (BMS).

 

The BMS monitors the voltage of individual battery modules connected in a series pack and controls the charging/discharging current to perform active and reactive power control by the battery pack. Various battery parameters such as state of charge (SoC) and depth of discharge (DoD) are calculated by sensing the voltage and current of the individual battery module. We chose the NI CompactRIO platform for implementation due to the high computing requirement for the BMS. A passive balancing algorithm maintains effective use of batteries and increases the life time of the battery pack.

 

 

A NI cRIO-9082 controller, an NI 9159 MXI-Express RIO chassis, a NI USB-8451 interface, a NI 9227 C Series module, NI 9225 C Series modules, and  NI 9485 modules increase the reliability, run time, and life span of the battery pack in the BMS. Figure 7 shows the connection diagram of the developed BMS. The NTU smart grid test- bed integrates a 5 kW lithium-ion battery pack through a bidirectional converter/inverter system (Figure 8). Eight NI 9225 modules connect to the NI cRIO-9082 system. One NI 9227 current sensor and three NI 9485 relay modules connect to the NI 9159 chassis. The individual battery modules communicate battery parameters through the NI USB-8451 interface using SMBus protocol. The Modbus TCP/IP communication protocol implemented on the FPGA module controls the bidirectional converter/inverter system. A Xilinx interface integrates the control algorithms we developed in LabVIEW into the FPGA module. Figure 9 shows the BMS algorithm implemented in LabVIEW.

 

Reduced Development Time, Easy Integration

We successfully developed a CEMS for the NTU smart grid test bed using NI CompactDAQ,  CompactRIO , and LabVIEW. The CEMS efficiently integrated the real-time solar power forecasting algorithm, improved MPPT algorithm for PV system, and robust BMS to deliver a reliable, cost-effective energy solution for smart grids. The CEMS scheduled the power flows to the loads between grid supply and DERs through control of the DER inverters. The developed smart grid test bed tested the applicability of the proposed energy solution to overcome the impact of intermittent energy generation of renewable energy resources. We used diverse, flexible products from NI to perform incremental research and easy integration with our existing system. Using the built-in driver hardware in LabVIEW, we significantly reduced development time for the CEMS software.

 

Author Information:

Sivaneasan Balakrishnam
Nanyang Technological University
50 Nanyang Avenue
639798
Singapore
Tel: 6567904795
Fax: 6567933318
sivaneasan@ntu,edu,sg

Figure 1. NTU Smart Grid Test Bed Architecture
Figure 2. CEMS Block Diagram
Figure 3. Solar Power Forecasting Algorithm
Figure 4.
Figure 5. Comparison of Measured and Simulated Solar PV Output Power
Figure 6. (a) PV Power Versus Voltage and Current Versus Voltage Characteristics (b) MPPT Tracking by EPP Method.
Figure 7. NI Hardware Integration for BMS
Figure 8. BMS in LabVIEW (a) VI Diagram (b) GUI