Design Of Fire Monitoring And Warning System Of Photov Oltaic Confluence Box Based On 4G

Solar energy is one of the most effective types of energy to solve the energy crisis and environmental pollution,so the photovoltaic industry has developed rapidly.The combiner box is a direct key equipment connecting the photovoltaic array and the inverter,so whether the photovoltaic combiner box can operate normally directly affects the normal operation of the photovoltaic system.Therefore,it is essential to monitor the environment in the photovoltaic combiner box and to warn of the fire that may occur in the box.Traditional fire early warning methods include threshold method,trend method,etc.,but these methods usually only consider the change of one characteristic parameter,and when the fire environment becomes complex,false alarms and understats often occur.With the development of artificial intelligence technology,intelligent algorithms have gradually been applied to fire early warning,and these algorithms have strong capabilities in the identification of fire conditions.Therefore,it is of great significance to design a fire early warning system with highaccuracy.The main contents of the photovoltaic combiner box fire early warning system designed in this paper include:First,the current development status of the combiner box and fire early warning technology at home and abroad is summarized,and the photovoltaic combiner box and fire early warning technology are introduced.On this basis,the fire monitoring requirements of the photovoltaic combiner box are analyzed,and the principle of fire generation and the development process are introduced,and the multi-data fusion technology is also introduced,which lays the foundation for the selection and data fusion of the three fire early warning parameters in this paper.Second,the software and hardware design of the combiner box fire early warning system was completed.The system takes the high-performance STM32F407ZET6 chip as the core control chip,the high-sensitivity MQ-7,MQ-2 and DS18B20 sensors as the main acquisition devices,and the farther EC20 module as the main wireless communication module,and designs the acquisition module,the main control module,the 4G communication module,the alarm module and other hardware modules.At the same time,the software of the above hardware modules is designed.Third,the Nuclear Limit Learning Machine(KELM),Particle Swarm Algorithm(PSO),Whale Algorithm(WOA)and Harris Eagle Algorithm(HHO)areintroduced,andthe PSO-KELM,WOA-KELM and HHO-KELM fire warning algorithms are designed.Through the comparative analysis of the selected simulations of three fire early warning algorithms,the HPO-KELM fire early warning algorithm has a high accuracy and strong generalization ability.Finally,because the HHO algorithm has problems such as slow convergence,lack of diversity in initial populations,and easy to fall into local optimal solutions.Therefore,the HHO algorithm is optimized by reverse learning,differential evolution algorithm and quadratic difference method.Immediately after the use of the improved Harris Eagle algorithm to optimize the nuclear limit learning machine,through simulation analysis and comparison of the improved Harris Eagle optimized nuclear limit learning machine,the early warning accuracy rate reached 98.3%,the convergence speed is faster.Figure [62] Table [9] Reference [69]...