Ⅰ. Introduction

To protect the environmental conditions, gradually need to replace renewable energy sources for power generation since coal, water, etc. are getting exhausted and the energy distribution become an man kind activity to use safely, cleanly and sustainably is a challenge in that 21st century and future. The Photovoltaic (PV) is currently one of the industry's leading solutions for high energy density, and sustainability using solar energy.

A smart green city that uses the renewable energy is a trend for humanity. To use the solar energy efficiently, it is important to monitor the power output from the solar panel and enable the way to output maximum power all the time. This facilitates to restore energy output efficiently from the solar panels. If the solar panels fails output the maximum power output, that may be due to physical connection problems or dust accumulation on the solar plates and less sunlight on the solar panels affects the performance solar. The solar energy management is a challenge task because solar energy break the conventional methods of planning due to their energy fluctuations over time periods. The solar energy fluctuations due to sunrise and sunset, solar energy output conversion ration in the panel direction with sunlights, and due to changing weather conditions. This paper proposes the design of PV array connected to resistive load through DC-DC boost converter with perturb and Maximum Power Point Tracking (MPPT) controller to enable maximum power output from solar panels. The proposed system simulated using MATLAB and results are discussed under constant as well as variable irradiation and temperature conditions.

The Wireless sensor network (WSN) is increasingly used to improve applications such as military monitoring, medicine, transportation, environmental conservation, agriculture, family health care and control. Industrial processes, among other applications. There are many lower energy sensors is deployed for WSN in different environments to collect, process, analyze and monitor elements in real time due the emergence of Internet of Things (IoT),

To establish a smart green energy city, its mandatory to include the monitoring devices that measure module-level parameters placed on different locations spread across the city wherever the solar panels are installed and that can provide detailed module-level measurements to create an accurate smart city energy model (Rusydi et al., 2016). However, powering the sensor and control devices becomes a challenge tasks because the sensor and control devices are works on batteries (Climent et al., 2016).

The battery system used in IoT sensor end device includes the main battery (non-rechargeable battery) and extra battery (rechargeable battery). The main battery is used as an energy source for sensor end devices and the lifetime of these sensor end devices is the time to discharge below the minimum charge required by the sensors (Escolar et al., 2013). Although the batteries that used sensor end devices have a high energy density with limited life time. For long-term deployment of sensor devices, batteries need to be replaced regularly and the battery replacement can be expensive and not feasible in the case of remote environmental monitoring (Penella and Gasulla, 2007;Penella et al., 2009). And the limited energy based sensor devices also has some important performance parameters in wireless sensor network systems, such as sampling frequency and maximum transmission distance between sensor nodes, etc.(Capella et al.,, 2013). The sensor devices with higher sampling frequency consumes more energy and increases the transmission power.

These days, there are different ways used in energy harvesting systems for power supply: super capacitors, batteries and a combination of both (Vracar et al., 2016). To achieve long-term operating condition, harvesting secondary energy batteries (or super capacitors) as an alternative storage element is essential for sensor devices used in smart cities .

In this paper, we propose an automated solar monitoring system based on IoT that allows automatic solar monitoring from anywhere via the internet using LoRaWAN network. This makes remote monitoring of solar plants very easy and ensures the best power output.

Ⅱ. Wireless Sensor Network

IoT based wireless sensor network solution is chosen for smart cities because the transfer data between the monitoring equipment to the city center control room is not feasible through the wired network solutions, thus the data cable need to be use very long, takes a lot of installation time, and the costs for wires and installation are very high. There need to be install new cables when extent more monitoring devices or if the monitoring device is moved to other locations in the city.

There are a few wireless network based key factors to be investigated when evaluating wireless technology to be used in smart cities monitoring. The key factor includes: scope, topology, data requirements, network security and energy needs. The importance and weight of different wireless networks factors depends on the specific application in which wireless technology is deployed for smart cities remote monitoring (Ko et al., 2010;Himanshu et al., 2017).

The required setting for the WSN is one in which sensor devices for monitoring need to be placed on different locations spread across the city. The required number of monitoring devices throughout the city is depending on the area of the city and the accuracy level of measurement details required. The monitoring sensor devices measure the module-level parameters such as current, voltage, rear temperature, ambient temperature and radiation. The measured datas are then sent over the wireless network to a central port to be viewed , stored, and analyzed.

Bluetooth, ZigBee, Wi-Fi, Cellular, Sigfox and LoRa are the most used wireless technologies for WSN and the deployments and public relations (PR) profiles according to system requirements. LoRa is chosen as the wireless technology for the WSN because LoRa has a very long range, LoRa can be configured to operate as a star or mesh network, LoRa has different data rates that is sufficient in transmitting measurement data across the wireless network at the required intervals and LoRa is a low power technology that use little power.

Ⅲ. Proposed System

The proposed system uses the low power micro-controller for data acquisition to analyze signals from sensors such as current sensors, voltage sensors, solar radiation sensors and temperature sensors. The proposed system model block diagram is shown in <Fig. 1>. The total system to be monitored at the city control center via LoRaWAN. The LoRaWAN is low energy consumption wide area network based on LPWA. The smart cities abnormal power conditions and power failure will be promptly adjusted using proposed solution.

The amount of electricity generated from solar energy in part is provided directly to the DC load, such as the street lighting system, etc. The rest of power used to charge the batteries connected with public loads and then distributed to the connected grid.

Ⅳ. Implementation and Analysis

In this proposed system implementation, we used a wireless sensor node and gateway using DC-powered micro-controller (5V) to analyze and decode the signals received from the sensors.

The sensors are used to measure current, voltage, temperature and solar radiation. The analog signals from the sensor are fed into the micro-controller's analog port to decode the sensing data informations. The micro-controller estimates the sensing parameters according to the embedded sensing logical program. The sensed parameters are analysed and then transferred to the city center monitoring room via wireless network. The proposed system operation process is described in <Fig. 2>.

1) MPPT in PV Systems

The PV devices are semiconductor devices that are able to directly convert the incident solar radiation into electrical energy. The PV system response characteristics are shown in <Fig. 3>. We can be notice that at one particular voltage (Vmp) PV cell delivers maximum power (Pmax). The P-V (power-voltage) and I-V (current-voltage) characteristics change with and with change in irradiation and temperature. Hence, the voltage at maximum power point (Vmp), current at maximum power point (Imp), open circuit voltage (Voc) and short circuit current (Isc) change with change in irradiation and temperature.

In the P-V curve <Fig. 3(b)>, there is a peak voltage point called maximum power point (MPP) which always occurs in the knee of the curve (Alex et al., 2012), where the generated PV power is maximized as shown in <Fig. 3(b)>. In the maximum power point tracking (MPPT) algorithms, searches the MPP by comparing the output power of the PV module before, and after the duty cycle of the converter is changed.

The MPPT controller measures the current values I and V, then calculate the deviation ΔP, ΔV and check:

Then update the new values in place of the previous value of V, P and then measures the parameters I and V for the next working cycle. Therefore, need to track maximum power point (MPP) to get optimum power regularly from solar panels. There are a lot of MPPT techniques been used in solar panels. The most commonly used MPPT techniques are incremental conductance and perturb and observe MPPT. In this proposed system, MPPT with perturb and observe (P&O) technique is used to get maximum power from solar panels. The proposed system is modeled with Battery charging system using MATLAB and performance analysed according to the MPPT controller with P&O technique as shown in <Fig. 4>.

The proposed system simulation considered 12V voltage battery for charging under constant irradiation (1000W/m2) and Temperature condition (25℃). The Battery voltage charged gradually to reach 12V threshold and then stops. The battery capacity increase over time and reach threshold value as shown in <Fig. 5>. Initially PV is made to operate at duty cycle equals to zero for 0.1 sec and step size (ΔD) are taken as 0.5. In <Fig. 5>, we can be observed from PV array characteristics that the PV panel delivers maximum power is 285W at Voc is 37V and Current of Isc is 8.6 A.

To confirm the proposed concept efficiency, we compare the proposed system performance with P&O MPPT algorithm and without P&O MPPT algorithm.

The simulation was first run without MPPT mode, bypassing the MPPT algorithm block in the circuit. It is observed that when we do not use an MPPT algorithm, the PV panel output panel continued to generate around 165 Watts and Power output of DC-DC boost converter around 149 watts for a solar irradiation value of 1Kilowatts per square meters.

Then the simulation was then run with MPPT mode by using the MPPT block in the circuit and the controller was fed the Pref as calculated by the algorithm. It is observed that with the same irradiation conditions, the PV panel output panel continued to generate around 280 Watts and Power output of DC-DC boost converter around 260 watts.

The proposed solar panel monitoring system emulated using Solar Panel Model : IT-100W (21.5V), Rocket ES 15-12 (12V, 15AH), Custom DC-DC Controller with P&O MPPT controller, and Arduino Open source Hardware controller board interfaced with current, voltage, temperature, and light sensors and LoRaWAN controller as shown in <Fig. 6>.

In this emulation, the proposed LoRaWAN network reliability for solar panel integrated smart city power system monitoring evaluated using packer delivery ratio (PDR) matrix. The PDR matrix is computed based on 500 packets exchanged between end-to-end node that is from the city monitoring center sink node to the each node of the solar panel integrated sensing nodes in real-time. We observed that the average PDR of 98.73% for the LoRaWAN with varying inter-packet intervals in real-time and this proposed observation confirms that LoRaWAN is suitable wireless connectivity for smart city power utility monitoring.

Ⅴ. Conclusion

In this paper design and simulation details of PV array connected to resistive load through DC-DC boost converter with P&O MPPT is presented and efficiency of overall system is found to be 93%. The implementation results are discussed under constant as well as variable irradiation and varying temperature conditions. This paper found that Perturb and Observe MPPT technique is easy to implement and shows good performance under steady change in atmospheric conditions but is less suitable under fast changing atmospheric conditions. The real-time emulation confirms that the LoRaWAN network works with the average PDR of 98.73 with varying packer size and inter-packet intervals and confirms that LoRaWAN can provide efficient power utility monitoring for smart cities.

This paper is a new development direction for providing green energy source to smart cities as an alternative power sources for generating electricity are increasingly depleted. This analytical paper provides optimal way to reduce energy costs when deploying the system at the same time create an optimal and intelligent operation system that is easy to put into practice. A smart city can not only self-supply electricity, but also replenish the lack of electricity for grid.