Design of Intelligent Meter Reading System in Green Sensor Network

　　introduction 

　　With the increasing pressure on resources and environment, low-carbon economy and energy conservation and emission reduction have become the hot topics of research and attention recently. Focusing on the macro-goals of resource conservation and environmental friendliness, "green channel" and wireless sensor network (WSN) have become a feasible choice. WSN is a task-oriented wireless self-organizing network composed of multiple nodes. It integrates sensor technology, embedded computing technology, modern network and wireless communication technology, distributed information processing technology, etc. It can monitor, sense and collect information of various environments or monitoring objects in real time through various integrated micro sensors, process the information through embedded systems, and transmit the sensed information to user terminals through random self-organizing wireless communication networks in a multi-hop relay manner. "Green communication" mainly adopts innovative high-efficiency power amplifier, multi-carrier, distributed, intelligent temperature control and other technologies, and cooperates with flexible site scenario models to actively transform base stations to achieve the purpose of reducing energy consumption. Green sensor network (GSN) that integrates green communication and WSN technology can effectively reduce node energy consumption, which is conducive to extending the service life of sensor nodes and reducing other environmental problems caused by frequent battery replacement or disposal. Due to these advantages, green sensor networks have broad application prospects in smart electricity consumption, building sensing, and environmental monitoring systems. 

　　1 AMI in Green Sensor Networks 

　　The emerging smart grid technology covers all aspects of power generation, transmission, distribution and consumption, and puts forward new requirements for the real-time, reliability and power consumption of communication. Applying green sensor network technology to smart power consumption, on the one hand, is to design algorithms and network protocols at all levels in the communication construction of smart power consumption, including optimization and comprehensive consideration from the physical layer, channel coding, media access control layer to routing, transmission, and application layer to reduce system power consumption and electromagnetic radiation; on the other hand, it is to achieve better measurement reliability, stability and energy saving by optimizing the measurement equipment in smart power consumption. As the basic link of the intelligent power consumption end, the automatic meter reading (AMR) in smart power consumption has begun to transition to the advanced measurement system (AMI). The realization of AMI has become the first step in the realization of smart grid technology . The general AMI system structure is shown in Figure 1.

　　In terms of equipment, AMI mainly includes: controllable electrical appliances, smart meters , collectors, concentrators, data processing centers and various meter reading terminals. In terms of communication networks, AMI includes the user's indoor network (HAN), the network between smart meters and collectors, the communication network (LAN) between collectors and concentrators, the network (WAN) between data concentrators and data processing centers, and the network for clients to access data processing centers. As the basic equipment for realizing AMI, the design, production and use of smart meters have received widespread attention in the industry. In November 2009, China State Grid Corporation issued a new standard for smart meters and held a centralized bidding for more than 20 million smart meters for the first time in 2010. In the next few years, China plans to install 130 million smart meters, and the total investment in smart meter reading will reach 38 billion yuan. 

　　2 Design of the electric meter 

　　The hardware system design of smart meters is divided into several main modules, including: metering module, processor module, RTC clock module, display module, storage module, communication module and power module. The communication module involves the selection of multiple communication methods such as RS485, power line carrier, short-distance wireless communication, etc. The system design block diagram of the entire meter is shown in Figure 2.

  　　2.1 Metering module 

　　The metering module is one of the core modules of the smart meter . It calculates the signals obtained from current sampling and voltage sampling to obtain energy equivalent pulses, power quality parameters, etc. The metering chip uses STPM01. The front end of the chip integrates analog current and voltage sampling, amplification, filtering, amplitude, and phase compensation units, and the back end is a DSP processing unit and SPI interface. It can calculate active power, reactive power, apparent power, grid frequency, voltage effective value and instantaneous value, and current effective value and instantaneous value. The structure of the metering module is shown in Figure 3.

　　STPM01 and the processor form a master-slave mode metering solution. The signal passes through the SPI port and is connected to the SPI pin of the MCU through the ADUM1411 four-channel isolator. The MCU sends the initialization and calibration signals to the metering module and modifies its configuration bits APL, KMOT, MDIV, TMP, etc. STPM01 sends the configuration status information and calculated measurement data to the MCU module. Here, the APL bit is set to 0, so that the voltage zero-crossing signal is output at the MOP pin and the watchdog signal is output at the MON pin. Configure the KMOT bit to output an apparent power pulse of 3000Pulse/kWh after optocoupler isolation. When calibrating the meter, the MCU writes the preset calibration data to the 56-bit OTP memory of the metering chip. If modification is required, the parameter value can be adjusted in the processor module and re-written.

{$page$}
　　2.2 MCU module 

　　The processor used is STM32F103, which is a 32-bit microprocessor based on the Cortex-M3 core and has 64 pins. The processor operates at 72MHz and has a built-in 128K
bytes of Flash memory and 20K
bytes of SRAM. A 7-way general DMA can be used to directly manage data transfers from memory to memory, device to memory, and memory to device. The MCU module structure diagram of the meter is shown in Figure 4.

　　The module is based on the processor, receives the status signal and energy information sent by STMP01 from the SPI port, expands the EEPROM storage device M24512R and the RTC clock M41T83 through the I2C port, and connects to RS485 communication through USART. When driving the LCD module, in addition to the common initialization pin RESET, 6 processor pins are used as control terminals, of which PB2 is used as the LCD backlight control terminal, and PB8 is the timer pin, which is used as the LCD signal interrupt request. In addition, according to
the characteristics of STM32, when the two-bit values ​​of BOOT0 and BOOT1 are set to 0X, the mode is to read the main flash memory; when set to 11, the mode is to read the built-in SRAM. [page]

　　2.3 Communication and interface card module

　　The scheme designs three communication modes: RS485, power line carrier, and infrared. The power line carrier uses the ST7570 integrated carrier communication chip. The interface designs include ESAM card, Smart
Card, and miniUSB interface. The circuit diagram of infrared communication is shown in Figure 5.

　　2.4 Power Module 

　　The power supply adopts a switching power supply structure, and its design schematic diagram is shown in Figure 6. The mains power is led into a high-frequency transformer after lightning strike, overcurrent protection, overvoltage protection, filtering, and rectification. It is divided into three windings from the secondary side of the transformer and led out after being stabilized by the KF50B voltage regulator. One way supplies power to the RS485 communication part; one way is divided into two branches, one +12V directly supplies power to the relay and power carrier, and the other is converted to power the MCU and wake-up circuit after DC-DC conversion; the third way supplies power to the isolation, comparator and other modules. The control switch part adopts the VIPER27 chip, which integrates a current PWM switch and an N-channel MOSFET with a minimum breakdown voltage of 800V. The secondary side is divided into three windings, which increases the wiring difficulty to a certain extent, but simplifies the isolation between circuit modules. To reduce electromagnetic interference, a 330pF/100V capacitor is connected to the output +12V winding, a 330pF/2kV capacitor is connected to the +5V output winding, and a separate anti-crosstalk capacitor is also connected between PG and ground.

 　　2.5 Software Design 

　　Combined with the underlying driver of the metering chip, the software design of the meter is divided into modules. The software mainly consists of the initialization and system management main program, the clock module program, the display module program, the power management program, the communication module program and the event alarm program. The power line carrier and infrared in the communication communicate according to the established protocol of the electricity department. The event alarm program monitors the meter overload, electricity theft and cover opening events. 

　　3 Wireless HART and Meter Reading Communication
{$page$}

　　There are many protocols at all levels of the green sensor network, such as Direct Diffusion, LEACH, S-MAC, ZigBee, etc., which can provide abundant redundant paths with the topology structure, improve the reliability of data transmission, and enhance the network's ability to resist environmental interference. With the development of AMI technology, the use of green sensor networks for smart meter reading will be a new trend, but most wireless meter reading is based on private communication protocols, while wirelessHART is built on HART and is the most widely used international standard in the current industry. The comparison between this protocol and ZigBee is shown in Table 1. WirelessHART has higher reliability, security and lower device power consumption than ZigBee. In this design, the collector and concentrator use STM32F103 processor and CC2520 transceiver chip. The concentrator adds a GPRS module as a remote communication channel for meter reading. The schematic diagram of the meter reading network structure is shown in Figure 7.
Each collector hangs 16 smart meter units and has routing functions. The gateway provides an interface for the collector's field equipment and the management host station, collects the collector's meter data through the wirelessHART wireless network, and uploads the data to the power department's application management host through GPRS.

　　4. Detection 

　　After testing, the single-phase smart meter has an operating voltage of 220VAC ± 20%, a frequency range of 50Hz ± 10%, an accuracy of 0.5, Ib = 5A, Imax = 60A; it can realize four-quadrant energy measurement, voltage, current parameters, power factor measurement and display; fast digital calibration and single-line tampering detection; programmable energy pulse LED output; multi-rate, prepaid account management and other functions. AC power supply and battery switching, power consumption is within 3W in AC mode, less than 6.5mA in battery mode, and less than 52μA in standby mode. The meter is connected to the wirelessHART protocol based collector and concentrator wireless network meter reading through RS485, and tested under experimental conditions. The transmitting node power is controlled within 50mW, the communication distance is within 200m, the one-time collection success rate and the periodic collection success rate are both above 99%, and the meter operates normally. 

　　5 Conclusion 

　　This paper applies the technical concept of green sensor network to smart electricity consumption and AMI, and designs a smart meter that meets the requirements. The meter uses STM32 processor, and its operating efficiency is better than that of 16-bit solution, while the power consumption does not increase much. In the implementation of green sensor network meter reading, the smart meter is combined with the collector as the network node in the sensor network, and the wirelessHART protocol is used to build a local communication network for meter reading, realizing the function of the meter and the smart meter reading system. Compared with the ZigBee protocol, the meter reading system has certain improvements in node power consumption, reliability and security. Its characteristics are also that the main chips for building the meter and the system are basically within the product framework of ST company, which simplifies the maintenance and upgrade of hardware. At present, the smart meter solution has been basically determined, and the potential of green sensor network technology in smart electricity consumption, such as further reducing system power consumption, improving AMI's anti-electromagnetic interference ability and communication efficiency, and improving service quality, still needs to be further explored and studied.